Shift work and irregular working hours

Autistic Mirror (outside view for the surrounding people, employer role)

Behind this is neurology, not will.

Predictive coding describes how the brain constantly builds predictions about its environment and compares them with the actual input. In autistic people this predictive model is particularly stable and precisely calibrated. That sounds like an advantage at first, but it has a direct downside: when the sleep-wake rhythm, the daily structure or the working hours change abruptly, the entire predictive model has to be rebuilt. That is not a quick switch, but an energetically expensive process that takes hours to days. For neurotypical colleagues the same process runs faster, because their model responds more flexibly to deviations.

A second factor adds to this: many autistic people have a shifted or irregular circadian rhythm. That is not a preference but a physiological characteristic of the internal clock. Night shifts then do not meet a flexible system, but a nervous system that is already biologically timed differently. The exhaustion after night shifts in this constellation is not an expression of a lack of effort, but of genuine physical desynchronisation.

Autistic inertia is a third mechanism. A state that has been started, whether a sleep rhythm, a work routine or a bodily state, is neurologically hard to switch. Short-notice shift changes are therefore particularly costly: the old routine is still active, the new one does not yet have a stable basis. In this transitional state cognitive capacity is markedly reduced, and that is exactly why mistakes occur. That is not a lack of concentration, but exhausted processing capacity.

The mistakes are therefore a direct signal from the nervous system, not warning signs of missing motivation.

What helps here is not admonishment, but discussing shift planning together with this team member: stable core hours, as few short-notice changes as possible, and where night shifts are unavoidable, planning enough recovery time afterwards. The person themselves knows best which structures work.

Autistic Mirror (outside view for the surrounding people, employer role)

Behind what you are observing lie two different mechanisms that overlap. It makes sense to look at them separately before looking at how they interact.

The first mechanism is autistic inertia. Being in a state once, that is in a rhythm, a routine, an established daily flow, does not simply mean habit in the everyday sense for many autistic people. Switching between states is neurologically expensive, because basal ganglia and resting-state circuits are wired differently. That means: a swap from early to late shift is not just a different time. It is a complete system change that recalibrates the time of waking, meals, light conditions, social contacts at the edges of the day and the internal timing of an entire day. This rebuild costs capacity, regardless of how many hours are worked.

The second mechanism is intolerance of uncertainty, which runs directly through predictive coding. The brain constantly compares expectations with what actually arrives. The more precise the established predictions were, the more expensive every deviation is, because the whole model has to be reweighted. Predictive coding is the computational mechanism: the brain internally sends ahead what is about to happen and compares that with the actual input. The more detailed this model, the more capacity an unexpected update costs. A short-notice shift swap is a prediction error on several levels at once: when it gets light, when the bus comes, when the body signals hunger, when the energy for certain tasks is there. Each of these small errors costs something, even if none of them sounds big.

These two mechanisms amplify each other. Inertia makes the switch itself expensive. The prediction costs make the adjustment phase afterwards expensive. That does not add up linearly, because the energy for adjustment is the same energy that also carries the work. The total number of hours stays the same, but the baseline load under which those hours are worked has shifted completely.

There is also a third level which, while not an independent autistic mechanism, clearly worsens the situation: short notice means that the predictive model could not be built in advance. Long-term predictability is not a comfort issue for autistic people, but a condition for the system to work in a stable operating mode at all. The shorter the notice, the higher the price.

What this means for planning: shift swaps that count as uncomplicated for neurotypical employees because the hours stay the same are structurally more demanding for autistic employees, regardless of individual flexibility or goodwill. Early notice and a say in planning are not special treatment here, but accessibility.

Autistic Mirror (outside view for the surrounding people, employer role)

Night duty usually delivers fragmented, shallower sleep, because the circadian rhythm and shift times run against each other. The HPA axis, the control loop of hypothalamus, pituitary and adrenal gland, produces higher baseline cortisol levels and recovers more slowly. That is not abstract: cortisol dampens prefrontal activity among other things, and the prefrontal cortex is exactly the area that controls sensory filtering when automatic pre-filtering is not enough.

This is the core. Sensory gating is a largely preconscious process that filters out redundant or irrelevant stimuli before they even reach conscious perception. In many autistic people this sieve is more permeable from the outset than in neurotypical people. That means: more raw sensory material arrives at the cortical level and has to be laboriously reprocessed there. What neurotypical colleagues tune out automatically requires active, capacity-intensive processing in autistic nurses.

Poor sleep hits exactly this system as a precondition. Sleep architecture, that is the structure of the sleep phases with their REM portions and the nightly melatonin-cortisol shift, determines how much processing capacity the following day provides at all. After night duty this budget is already reduced before the first beep of the monitor, not through stress, but through the neurobiological state of the nervous system. The HPA axis has not gone through a normal recovery arc.

The ward noise, that is monitors, patient calls, doors, colleagues talking in the background, now meets a sieve that is already more permeable, and a prefrontal cortex whose reprocessing performance is further throttled by sleep deficit and raised cortisol. The result is sensory processing fatigue earlier and with a lower trigger threshold. What would normally lead to overwhelm only after several hours on duty can be reached in the first hour after night duty.

That is not a sensitivity problem and not a sign of a lack of resilience. It is the predictable result of an interaction between a neurobiologically based gating profile and a sleep-related lowered starting budget.