Exam Wellbeing: Sleep, Stress and Performance

The final fortnight before MRCS Part B is when a particular logic tends to take hold: every hour of sleep starts to look like an hour of revision sacrificed, every meal becomes a delay, exercise feels indulgent. The cognitive science suggests the opposite. A candidate who walks into the examination rested, calm, and properly nourished will, on average, outperform a candidate who walks in with two more hours of revision and four fewer hours of sleep. None of this is a moral position; it is performance physiology.

What follows is an evidence-led wellbeing protocol for the four weeks before the examination — treated not as self-care but as a set of marginal gains with credible effect sizes for cognitive performance, mood, and recall.

The Sleep Finding That Matters Most

The most directly relevant piece of evidence is also one of the simplest. Galland and colleagues studied 216 medical students sitting a high-stakes OSCE, recording the sleep they had in the month before the exam and the night before, then comparing those data with the marks the College awarded [1]. Sleep duration the night before the OSCE was significantly associated with score (p = 0.026). Students who achieved a distinction grade were significantly more likely to have slept more than six hours the night before. Almost 40% of the cohort had less than six hours that night.

The authors are careful about causation — better-prepared students may also have been less anxious, and therefore slept better — but the correlation is reproducible and clinically meaningful. The students at the top of the mark distribution were the rested ones.

A second finding is worth mentioning briefly because it explains why sleep matters technically. Memory consolidation — the process that converts the evening’s revision into the next morning’s recall — takes place largely during NREM sleep, mediated by thalamocortical sleep spindles and slow oscillations [2,3]. The candidate who studies hard and sleeps badly is encoding new material into a system that is then unable to file it. The reading does not become recall without the night that follows.

The Stress Curve, and What Most Candidates Get Wrong About It

The dominant cultural narrative about exam stress is binary: stress is bad; calm is good. The evidence is more useful than that.

McEwen’s work on allostatic load distinguishes acute stress responses (sympathetic activation, raised cortisol, focused attention) from chronic stress states (prolonged cortisol elevation, prefrontal degradation, impaired working memory) [4]. The former is performance-enhancing within a range. The latter is performance-destroying. Folkman and Lazarus showed that the same physiological state can be appraised as a threat (which impairs performance) or as a challenge (which improves it) — and that this appraisal is modifiable [5].

This matters operationally. The candidate trying to be “calm” before the examination is sometimes pursuing the wrong physiological state. Mild sympathetic activation — the focused alertness of a well-managed pre-station moment — is performance-enhancing. The intervention is not to eliminate arousal. It is to calibrate it.

Does anxiety actually cause failure?

The clinical instinct is to assume that the anxious candidate fails. Sturpe’s 2020 systematic review of OSCE anxiety challenged this directly. Across eight studies examining the relationship between measured anxiety levels and OSCE performance, six found no statistically significant association [6]. Anxiety, in isolation, does not predict failure. What predicts failure is what the anxious candidate does with the anxiety — whether they freeze, ruminate, or convert it into productive arousal.

The implication is liberating. The candidate who walks in with palpable anxiety should not interpret that as a predictor of failure. The body is doing what bodies do under high-stakes performance conditions. The work is to keep moving through the station regardless — not to wait for the anxiety to subside before performing.

Recognising Chronic Stress — Before It Becomes Something Else

Acute exam stress is one thing. Chronic stress — the kind that builds quietly over a CT2 year of busy rotations, on-calls, and revision — is another. It is also, in the trainee population, common, under-recognised, and potentially serious. This section is the most important one in this article.

McEwen described allostatic load as the cumulative wear-and-tear that occurs when the stress response system remains chronically activated rather than turning off cleanly between stressors [4]. In healthcare workers specifically, chronic activation is associated with hypocortisolemic profiles, blunted morning cortisol, and the constellation of symptoms that maps onto burnout [7,8]. The pattern is not always dramatic. It is often a gradual loss of energy, motivation, and resilience — mistaken at first for being tired, then for being a bit low, then for something more.

What helps, in the evidence, is also straightforward and worth listing plainly. Regular sleep that respects the diurnal rhythm. Aerobic exercise three times a week. Social contact maintained, not deferred. Sustained time off (not just rest days that are filled with revision). Caffeine and alcohol kept inside sensible limits. And — the one most trainees defer — talking to someone early, before the picture deteriorates. The recovery curve is steep when intervention is early, and shallower when it isn’t.

Exercise as Free Dopamine

One of the cleanest interventions in the literature is also one of the easiest to misuse. Aerobic exercise, performed regularly and at moderate intensity, has a measurable downstream effect on baseline cortisol, on the HPA axis’s reactivity to subsequent stressors, and on the dopaminergic systems that support motivation and concentration [9,10,11].

The 2025 network meta-analysis of exercise modalities for cortisol reduction (Liu et al.) identified moderate-dose continuous aerobic exercise as among the most effective interventions: regular moderate aerobic sessions restored diurnal cortisol rhythms and attenuated chronic stress reactivity, with the dose-response curve flattening at higher intensities [11]. Random higher-intensity sessions transiently raise cortisol; the chronic adaptation comes from consistency, not from any single hard workout.

Subjectively, what trainees notice first is mood and motivation. Salamone and Correa’s work on mesolimbic dopamine showed that aerobic exercise reliably increases striatal dopamine release in the hours afterwards [12]. The effect is small per session, but cumulative. It is, in effect, free dopamine — available daily, with no withdrawal, no cost, and no contraindications for most trainees.

The Weekly Social Treat — Not the Sugar One

Most candidates plan their week around revision blocks. Few plan it around recovery. The literature on burnout in healthcare workers, and on workplace recovery more broadly, is consistent: protected social contact — particularly with people outside the candidate’s immediate professional circle — is associated with lower allostatic load, lower depressive symptoms, and better self-reported wellbeing [13,14].

One useful framing: protect at least one social commitment a week that is properly enjoyable and does not involve revising. A meal with friends. A film. A long walk with someone who isn’t a doctor. The temptation in the final fortnight is to cancel these; the temptation should be resisted. The marginal hour of revision gained by cancelling is small. The hour of sustained mood and motivation lost can be significant.

This is also the place to be explicit: the treat is not a sugary one. The post-revision biscuit / cake / chocolate habit feels rewarding but produces a measurable post-prandial energy slump and, in some candidates, evening sleep disruption from the late insulin response. The reward should be social or experiential, not glycaemic.

Diet and Supplements — an Honest Read of the Evidence

This section is shorter than candidates often expect, because the evidence base for dietary intervention in healthy young adults preparing for an exam is weaker than the wellness industry implies. The honest summary:

The honest position is this: diet matters in the way most other foundational behaviours matter. Eat regular, balanced meals. Drink water. Avoid the late-afternoon caffeine. Avoid the sugar spike. Beyond that, the evidence for cognitive supplementation in healthy young adults is thinner than the marketing suggests. The candidate looking for a 10% performance boost from a supplement is looking in the wrong place; that gain is more reliably found in sleep, exercise, and rehearsal.

The Four-Week Wellbeing Protocol

The 60-Second Pre-Station Routine

The corridor between stations is where one station’s outcome — good or bad — either stays where it belongs or follows the candidate into the next encounter. Three things, rehearsed in advance, make the difference.

The Underlying Principle

Wellbeing in the run-up to MRCS Part B is not a moral category. It is a performance category. The candidate who sleeps, who exercises, who manages caffeine timing, who calibrates their arousal, and who has rehearsed the corridor routine is not virtuous. They are well-prepared in a domain most candidates underweight.

The published evidence is consistent enough that the calculation is clear: a marginal hour of sleep, in the final two weeks, almost certainly returns more on exam performance than a marginal hour of revision. The same is true of a 30-minute walk, of a meal eaten on time, of a Friday evening with friends. The candidate who has revised exhaustively and slept badly has spent their preparation building knowledge they cannot then access. The candidate who has revised adequately and slept well has built knowledge they can deploy.

The work is to take this seriously enough to act on it — not in the final week, but four weeks out.