Sleep Continuity The Missing Variable Behind Your 3 PM Decision Fatigue

You can do everything “right” on paper—reasonable bedtime, 7 to 8 hours logged—and still hit the day feeling off. You start sharp, then the fog rolls in, focus gets jumpy, patience gets thinner, and decision fatigue shows up by mid-afternoon. If that’s your pattern, it’s usually not a discipline problem. It’s often a sleep continuity problem: sleep happened, but it didn’t stay asleep.

This article gives you clearer language for what’s going on, plus a way to test it without turning sleep into another performance project. You’ll learn how microarousals (brief, often unnoticed wake-ups) can drain recovery even when total hours look fine. You’ll also start thinking in three separate buckets: duration, continuity, and recovery yield, meaning what you can actually spend tomorrow in attention stability, emotional range, and decision patience.

We’ll cover why high-pressure work can fragment sleep without “stealing” hours (the hyperarousal trap), how it shows up at work (more variance, tone slips, sudden misses), and why wearables can help but also mislead. Then you’ll run a simple diagnostic: devices down at 9 pm. nothing else. Not as a moral routine—just a clean test.

Because “fine” isn’t a number on a sleep app. It’s your 3 p.m. decision quality. And if you’re ambitious, here’s the reframe I wish more people heard earlier: recovery is strategic resource management.

The “7.5 Hours but Wrecked” Problem: When Sleep Doesn’t Stay Asleep

If your day follows that wired → fog curve, continuity is the missing variable.

Public guidelines like the AASM/SRS ≥7-hour baseline matter (Watson et al., 2015). But sleep health is more than hours (Buysse, 2014). When the hours look fine but your day doesn’t, you’re missing vocabulary.

Microarousal debt: when sleep keeps rebooting

Microarousals are brief, often unnoticed awakenings: EEG-defined arousals where the brain shifts into a lighter, more activated state for ≥3 seconds, after ≥10 seconds of stable sleep (ASDA Atlas Task Force, 1992; AASM scoring rules). You can spend a full night in bed and still not get a clean restore.

Clinicians describe this with continuity terms like WASO (wake after sleep onset), sleep efficiency, and arousal index. These are not vanity metrics. They’re ways to name fragmentation.

Once you see microarousals as quick “reboots,” three things stop blending into one messy scoreboard:

• Duration: how many hours you got.
• Continuity: how often sleep breaks, even briefly.
• Recovery yield: what you can actually spend tomorrow, like attention stability, emotional range, and decision patience.

Research links sleep quality to performance (Pilcher & Huffcutt, 1996), and controlled fragmentation can degrade function—even when total sleep time is relatively preserved—by increasing vigilance lapses and reaction-time variability, the same “sudden misses” that show up at work (Bonnet & Stepanski paradigms). Hours are easy to count. Yield is what you feel. Continuity is the bridge between them.

Why “7–9 Hours” Became the Headline (and Why It’s Not Your Full Sleep Report)

Public guidance is duration-first because it’s simple and broadly true. Most adults do better at ≥7 hours (Watson et al., 2015). But when someone says, “I’m getting enough sleep and I still feel wrecked,” clinicians don’t stop at the clock. They look at maintenance insomnia, meaning difficulty staying asleep, which is part of DSM-5 insomnia criteria (APA, 2013).

If more metrics sounds like a trap, use one tougher question: did your sleep run as a stable block, or did it keep breaking and rebooting?

Then run a performance audit that doesn’t rely on vibes:

• By 3 p.m., is your attention steady, or does it swing between sharp and useless?
• When you make mistakes, are they slow-and-obvious, or sudden misses you “shouldn’t” make?

That instability shows up in vigilance research (Lim & Dinges, 2010). You don’t have to pick between ambition and sleep that actually restores.

The Hyperarousal Trap: Why Pressure Fragments Sleep Without Stealing Hours

The hyperarousal model of insomnia describes a common chain: sustained load raises baseline activation, threat-mode spills into the night, and sleep becomes lighter and easier to interrupt (Riemann et al., 2010). It’s physiology, not personality.

I learned this the hard way. In Stockholm, I kept telling myself I was “fine” because I was in bed for long enough. But the pattern didn’t lie: wired mornings, foggy afternoons, and a kind of brittle patience that made normal conversations feel like negotiations. Nothing looked dramatic on paper. My day did.

One layer that matters here—and that high performers often ignore because it sounds too “biology class”—is cortisol rhythm. Cortisol is supposed to peak in the morning (helping you wake and mobilize) and stay lower in the evening so sleep can deepen. Under sustained stress and late-night stimulation, that rhythm can get messy: you can wake feeling oddly revved, then pay for it later with flatter energy and shakier focus once the day’s demands stack up. Same hours, worse yield.

Wearables may hint at the pattern, like more interruptions or more “light sleep.” Some research also points to incomplete downshifting at night in insomnia cohorts (Spiegelhalder et al., 2011). In practical terms, that can look like “wired tired,” a sense that you never fully drop, or a night heart-rate trend that stays a bit higher than your own baseline. Caveat: heart rate and HRV are noisy. Alcohol, medication, training load, and individual differences matter. Treat wearables as a trend lens, not a diagnosis.

Open loops and the “on-call” brain

Modern work tools can train your brain to treat bedtime as a monitoring window. Unfinished work, unclear risks, and interpersonal uncertainty can make you feel psychologically on-call. You can be asleep while part of the system keeps scanning. In hyperarousal terms, that’s persistent activation during periods that should be quieter NREM sleep (Riemann et al., 2010).

Two common amplifiers can worsen fragmentation while preserving “hours”:

• Alcohol: it can knock you out early, then fragment the second half of the night (Ebrahim et al., 2013).
• Circadian misalignment (travel, shifted schedules): you may see earlier-than-usual wake time, more middle-of-the-night wakeups, and lower sleep efficiency even if your time in bed stays the same (Dijk & Czeisler).

If this is your mechanism, the daytime cost won’t always feel like classic sleepiness. It’s often variance: sharp moments next to sudden misses, the kind of instability your job punishes.

The Crash Report: How Fragmentation Shows Up at Work (and How I Triage It)

Sometimes it’s continuity. Sometimes it’s timing. Sometimes it’s alcohol. Sometimes it’s all three at once. My triage is simple: I look for what breaks first—attention, tone, or decision stability—and then I work backward from there.

The pattern is often: a fast start, then an unstable middle, then sharper edges.

Morning can feel awake but not settled. You’re scanning risk, checking the inbox, reprioritising, because the system never fully powered down (Riemann et al., 2010).

Later in the day, the first thing to break is often social calibration, not output: shorter messages, less tolerance for ambiguity, more reactive meetings. Vigilance is one of the most sleep-sensitive systems we have (Lim & Dinges, 2010). Sleep disruption is also linked to higher emotional reactivity and weaker top-down control, which is bad for tone, negotiation, and leadership presence. Meta-analytic evidence links sleep problems to poorer job performance and more counterproductive behaviour (Litwiller et al., 2017).

For example: it’s 2:40 p.m., you’re in a meeting you normally run calmly. Someone asks a fair question and you hear yourself answer too fast—slightly sharp. At 3:10 p.m., you approve a decision you already debated last week, then reverse it the next morning when you reread the thread and realize you missed one line. That’s not “you being dramatic.” That’s recovery yield leaking.

The Measurement Trap: When “Good Scores” Become Bad Decisions

Microarousals are EEG-defined events. Most consumer wearables infer sleep from movement (and sometimes heart rate). They’re often decent at detecting sleep, and often weaker at detecting quiet wake. That can underestimate awakenings and WASO (de Zambotti et al., 2018). A “solid” score can certify time still, while your brain is quietly logging reboots.

The opposite trap is orthosomnia: tracking turns sleep into surveillance, pressure goes up, threat-mode follows you to bed, and fragmentation gets worse (Baron et al., 2017). If tracking adds stress, it’s not measurement anymore. It’s a trigger.

If you’re the kind of person who likes data, keep it bounded for seven days:

• Track only bedtime and wake time (not sleep stages).
• Pick one daytime KPI: your 3 p.m. attention stability (1–5).
• Hide or ignore sleep-stage graphs and “readiness” scores during the week.
• Review once, at the end of the week, looking for direction—not perfection.

Re-anchor on outcomes: recovery yield over score-chasing

For one week, treat recovery yield as the KPI: steadier attention at 3 p.m., fewer tone slips, fewer decision reversals—outcomes that map to real job function (Litwiller et al., 2017).

Here’s a simple “yield dashboard” you can run without making this your new hobby:

• 3 p.m. attention stability (1–5)
• tone slips (count)
• decision reversals (count)

To test whether continuity-under-threat is your bottleneck, run one clean boundary: devices down at 9 pm. nothing else. This means phone, laptop, tablet, and TV off. What you do instead should be boring and low-light: a paper book, a shower, stretching, tomorrow’s to-do list on paper, lights dim. Then the next day, record your yield dashboard—especially what happens around 3 p.m. Screen exposure can delay circadian timing, and portable screens are linked to worse sleep outcomes (Chang et al., 2015; Carter et al., 2016).

If that doesn’t work, don’t spin into self-blame. Treat it like any other constraint problem: okay—so that didn’t move the needle. What’s the next constraint?

And yes, I used to say the same things you might be thinking now: “I’m already doing the basics,” “I don’t have time for rituals,” “this can’t be the thing.” Then I watched what pressure did to my nights and what it quietly did to my afternoons. It was never about being tougher. It was about getting sleep that stays asleep.

---

Run the test for 3–5 nights and let the afternoon tell you the truth.

• If devices-down improves 3 p.m. stability (fewer sudden misses, steadier tone, fewer reversals), continuity-under-threat is likely your bottleneck. Keep the boundary, and keep the yield dashboard small so you don’t turn sleep into another project.
• If it doesn’t change anything, widen the lens: maintenance insomnia patterns, alcohol timing, travel/circadian timing, or other sleep disorders can sit underneath the same “7–8 hours but wrecked” story. That’s when it’s worth discussing sleep maintenance with a clinician rather than trying to out-discipline it.

Because “fine” isn’t a number. It’s what you can reliably spend at 3 p.m. this week.