Vigilance systematically declines during night shifts

What it is

The circadian vigilance trough is a predictable dip in sustained attention, response speed, and judgment stability that occurs when the body's sleep-promoting circadian signal is strongest, deep in the biological night. It is not the same as feeling sleepy: someone inside the trough often rates themselves as fine while misses and response variability are already climbing — that gap between felt state and actual capability is exactly what makes the trough dangerous in safety-critical work.

Why it happens

The trough is not simply a byproduct of short sleep. Core body temperature and melatonin secretion both follow a roughly one-day cycle, and together they push cognitive alertness to its daily low point, typically somewhere in the second half of the night into the early morning. The deeper layer: even after a full, normal night of prior sleep, performance still degrades systematically while a person sits inside this physiological low — evidence that the trough is driven mainly by the endogenous circadian rhythm itself, not by accumulated sleep debt.

Homeostatic sleep pressure (which builds the longer someone stays awake) and the circadian trough are two independent curves that happen to stack: night shifts routinely collide with the peak of both at once, producing an effect larger than either alone. Brief high-arousal tasks, caffeine, or social interaction can mask the subjective feeling of sleepiness without touching the objectively measured lapses and microsleeps underneath — which is why self-report alone is a poor way to judge whether someone is safe to perform right now.

Studying it

To separate the circadian component from sleep debt, laboratories use constant routine protocols: keeping participants awake, recumbent, and under fixed posture and lighting for several days while tracking phase through markers such as core temperature or melatonin, alongside repeated psychomotor vigilance task (PVT) sampling. Field studies cannot isolate the variable this cleanly; they instead repeat measures of response variability, misses, eyelid-closure indices, and subjective sleepiness across shifts while logging sleep duration, light exposure, caffeine, and workload, using that pattern to roughly disentangle circadian phase from sleep debt.

Two confounds deserve caution: more senior staff are often assigned to day shifts, so a shift comparison can easily mistake a personnel difference for a shift effect; and a single sleepiness rating is not a substitute for a capability test, because felt sleepiness and actual response capability routinely dissociate inside the trough.

Where it stops holding

The trough's timing and depth vary by chronotype — morning types and evening types sit at different phases, so no single clock hour represents everyone. Long-term night workers partially adapt but rarely invert their rhythm completely; an occasional night shift carries a different risk profile than a standing one. This is also not an exclusively nocturnal phenomenon: day workers under severe acute sleep deprivation show a comparable degree of impairment, which shows that the real driver is the combination of sleep pressure and circadian phase, not whatever a roster happens to label the shift. Concrete scheduling limits and thresholds should follow sector regulation and occupational-health assessment rather than a single laboratory curve applied wholesale to a specific job.

Applying it

• Keep high-monitoring-load and irreversible operations outside the predicted trough where feasible; where it cannot be avoided, add independent verification or a second qualified check.
• Surface shift and time-awake information alongside task risk for scheduling decisions, without using it to judge or penalize an individual worker.
• Do not validate by asking "how tired are you": use real night-shift tasks or high-fidelity simulation with an embedded objective probe such as a vigilance task, comparing misses and response variability across the trough window; where available, slice historical error and near-miss logs by time of day to check whether they track the predicted trough.

Related

• Same group: Y6.05.2 Shift rotation direction affects circadian adaptation · Y6.05.3 Continuous work needs enforced rest limits · Y6.05.4 Interfaces should reduce nonessential load under fatigue
• Nearby: Y1.04 Vigilance decrement in prolonged monitoring · Y1.07 Shift handover
• Search terms: circadian trough · night-shift vigilance · sleep pressure