The Impact of COVID-19 on Anxiety in Chinese University Students

It's the middle of February in China, just two days before the new spring term is supposed to begin. Students across the country are at home — campuses are empty, classrooms are dark — as they log onto platforms they've barely used, figuring out how to start a semester through a screen. Then drop this number flat: fifteen percent of them, more than one in seven, are already in clinical anxiety territory. That's where this study begins. When the novel coronavirus emerged in Wuhan in December 2019, the Chinese government moved quickly. Cities went under quarantine in late January 2020, and a nationwide school closure followed almost immediately. The Ministry of Education didn’t simply pause the semester — it issued a mandate: "suspending classes without suspending learning." From mid-February onward, more than 100 million students shifted to online platforms overnight. For every undergraduate in the country, it was the first time in their lives that a new term had started this way. No lecture halls, no campus, no face-to-face anything. Chongying Wang and Hong Zhao saw an obvious but unanswered question at that moment: how anxious are these students, right now, and how does that compare to normal? To find out, they surveyed 3,611 undergraduates aged 18 to 24 from a top university in China, just two days before the spring term began. The timing was deliberate. This was the zero hour. The instrument they used is the Self-Rating Anxiety Scale, or SAS, developed by Zung in 1971. It consists of a 20-item self-report questionnaire covering four symptom domains — cognitive, autonomic, motor, and central nervous system symptoms. Each item is scored from one to four based on how often the symptom occurs, with five items reverse-scored. The resulting score runs from 20 to 80, and in the Chinese normative system, anything at or above 50 signals clinically relevant anxiety — mild at 50 to 59, moderate at 60 to 69, severe at 70 and above. Wang and Zhao distributed questionnaires through the university's online teaching platforms over a 48-hour window in mid-February, receiving 3,611 valid responses — a response rate of 95 percent. The sample included all four undergraduate years, all 26 colleges and schools, and students from regions across China. The female-to-male ratio was 1.48 to 1. Cronbach's alpha was 0.87, indicating solid internal consistency. Now for the numbers that matter. The mean SAS score across all 3,611 students was 40.53, with a standard deviation of 10.15. The Chinese national norm sits at 29.78. That's not a small gap — it's nearly eleven points, and the one-sample comparison yields a p-value below 0.001. This isn’t a pocket of unusually stressed students pulling the average up. It's a population-level shift. The entire distribution has moved. Within that elevated population, 557 students — 15.43 percent — crossed the clinical threshold of 50. Their mean score was 58.75. Breaking that anxious subgroup down: 358 students fell in the mild range, 143 in moderate, and 56 in severe. Here’s a historical comparison that sharpens the picture further. During the 2003 SARS outbreak, researchers documented a comparable student population with a mean SAS score of 41.93. The COVID-19 cohort's overall mean of 40.53 is statistically different from that SARS figure. More telling: among students who crossed the clinical cutoff during SARS, the mean was 55.14. During COVID-19, that same anxious subgroup averaged 58.75. The students who are most anxious now are more anxious than their counterparts were during the last major coronavirus outbreak. That difference is also statistically significant. So the sample as a whole is elevated, and the most affected students are more severely affected than precedent suggested. That's the core finding. Wang and Zhao then ask who within the sample is carrying the most anxiety. Female students scored higher than male students across the full sample, and that difference is significant at a p-value below 0.001. The raw means are close — 40.70 for females versus 40.28 for males — but with a sample this size, that gap is real. Within the clinically anxious subgroup, the pattern inverts in an interesting way: anxious males averaged 60.59 on the SAS, while anxious females averaged 57.55. So females are more likely to cross the anxiety threshold, but males who do cross it tend to score higher. Major field also mattered, but only among the anxious subgroup. Arts students who met the clinical cutoff averaged 59.09, compared to 58.34 for science students — a statistically significant difference at a p-value of 0.05. Among students below the threshold, no such difference appeared. Year of study produced differences across the full sample: first-year students differed significantly from second and third-years, and second and third-years differed from each other. Fourth-year students showed no significant difference from any group, but Wang and Zhao note that the fourth-year sample was only 116 people — too small to draw firm conclusions. Now the finding that arguably tells you the most. Wang and Zhao tested whether anxiety scores correlated with the number of confirmed COVID-19 cases in students' home cities and provinces. The answer was no — emphatically. Students from Hubei province, the epicenter, had a mean SAS of 40.07. Students from Wuhan itself averaged 40.51. Neither group differed significantly from the full sample. The Pearson correlations between SAS scores and regional case counts were effectively zero — t equals 0.03, degrees of freedom 110, p-value of 0.97 for cities, and similarly nonsignificant results for provinces. Geographic proximity to the outbreak predicted nothing about how anxious a student was. That's worth sitting with. You might expect that living in Wuhan — or coming from a province with tens of thousands of confirmed cases — would drive anxiety higher. It didn't. What this suggests, though Wang and Zhao are careful not to overclaim, is that anxiety during this outbreak wasn’t primarily about actual exposure risk. Information, perceived threat, and psychosocial context may matter more than whether you're physically near the hotspot. Everyone in China was receiving the same flood of news. The geography of the virus didn't map onto the geography of fear. The practical stakes connect directly to education. Wang and Zhao point out that 66.99 percent of students reported difficulties sitting still for extended periods, and 20.33 percent felt weak and easily tired. These aren't abstract symptoms — they're the conditions under which a student is supposed to sit at a desk and absorb a semester's worth of online lectures. The paper is direct about the implication: teachers delivering online classes need to account for this. Better mental health improves academic performance, and the two are not separable when an entire semester is being delivered through a screen to students who are also, in many cases, isolated at home. Wang and Zhao call for psychological support mechanisms to be woven into remote learning — structured health promotion, attention to sleep and diet and physical activity, and instructor awareness of student stress when designing online lessons. They also acknowledge the study's limits honestly. The sample comes from one university in northern China. The survey captured only anxiety, not depression or post-traumatic stress. It didn't collect information on whether students or their family members had been infected or where students were getting their information about the outbreak. And crucially, it's a snapshot. Wang and Zhao explicitly propose a follow-up study after students have lived through a full online term — to see whether anxiety levels change once the novelty of crisis-mode learning has settled into routine, or whether they deepen. What Wang and Zhao have given us is something specific and valuable: a baseline. This is the first measurement taken at the moment this generation of students pivoted into a new and entirely unprecedented form of education. The mean was 40.53 against a norm of 29.78. Fifteen percent were clinically anxious. The anxious were more anxious than SARS-era students had been. And none of it correlated with where you happened to live. You can't address what you haven't measured. This paper measured it. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.