How do sleep and test anxiety affect each other and academic performance, and through what pathways?

Why anxiety and bad sleep make each other worse

The most direct account comes from a study that followed 167 undergraduates in a statistics course during the two days before a midterm exam. A statistical model of the data showed a two-way relationship: worse sleep quality and restfulness went with higher test anxiety, and higher test anxiety went with worse sleep, with anxiety measured in the morning before the exam then predicting how the student performed [3]. The researchers describe this as a reciprocal process in which sleep and anxiety feed one another and jointly impair academic performance [3].

That framing treats test anxiety as something that unfolds over the days leading up to an exam rather than a fixed attitude toward tests, which is why the days right before the test are the interesting window [3]. The same study found that prior exam performance, being a non-native English speaker, and motivation in the class also predicted exam performance, so the sleep-anxiety loop is one contributor among several rather than the whole story [3].

Does REM sleep calm students down or wind them up?

A separate study tracked 52 college students who wore a mobile EEG device for a week during the semester, then rated their test anxiety, and rated it again just before finals. Habitual REM sleep turned out to be the strongest predictor of how much their test anxiety changed, but the direction was the opposite of what the researchers expected: more REM sleep went with a rise in anxiety rather than a drop [1].

The authors interpret this through a hypothesis that REM sleep adjusts activity in brain areas involved in encoding stress, making stress responses less sensitive overall but more selective [1]. In plain terms, the idea is that REM sleep may retune which things a student reacts to rather than simply turning the volume down. That is a proposed explanation, not a demonstrated purpose, and it sits alongside the finding that poor sleep in general correlates with more academic stress and lower performance among college students [1].

What the pattern looks like in real students

In medical students, a qualitative study described test anxiety as showing up with physical symptoms and sleep disturbances, and reported that these negatively affect academic work [2]. Because that study is qualitative, it documents how students experience and describe the problem rather than measuring the size of any effect, and it did not examine sleep as a separate factor in its analysis [2].

Among adolescents in Chinese public high schools, researchers asked whether sleep disturbance at night and sleep-related daytime impairment, such as fatigue, influenced academic performance and anxiety levels [4]. This extends the question beyond university students to younger learners, where the same ingredients, disturbed nights and tired days, are in play [4].

A study of university students examined performance anxiety and sleep problems together as predictors of academic achievement [5]. Read alongside the pre-exam diary study, the picture is consistent: anxiety and sleep problems each carry information about how a student will perform, and they tend to travel together [3][5].

Through what pathways does this reach grades?

The clearest pathway in this evidence runs through the exam morning itself: anxiety measured before the test predicted exam performance, and sleep quality was tied to that anxiety, so sleep appears to reach grades at least partly by shaping how anxious a student is on the day [3]. The researchers frame the combined effect as directly impairing performance while also mattering for overall student health [3].

A second, less settled pathway involves REM sleep's effect on how the brain handles stress. If REM sleep changes the sensitivity and selectivity of stress responses, as the recalibration hypothesis proposes, then habitual sleep patterns could gradually shift a student's anxiety trajectory over the weeks before finals [1]. That would make sleep a slow-acting influence on anxiety rather than only a night-before factor, but the evidence for the mechanism is interpretive and comes from a single small study [1].

What the evidence does not do is trace the full chain from sleep through anxiety to grades in one analysis. The pre-exam study modeled sleep, anxiety, and performance together but over a two-day window [3], while the REM study linked sleep to later anxiety without analyzing how that connects to academic performance [1]. So the loop is well supported at the level of anxiety and sleep, and the link from anxiety to exam scores is supported, but the complete pathway from sleep stage to grade remains assembled from separate pieces rather than measured end to end.

Sources used in this answer

Related research questions