COVID-19 symptom burden tracks with adolescent cognitive test performance
Adolescence is a period when the brain is still refining reasoning, language, memory, and motor control. Against that backdrop, a cohort of 3,860 adolescents in Lower Austria offers a concerning signal: COVID-19 symptom severity was associated with lower performance on several cognitive measures, even though infection history alone was not linked to worse results.
The pattern is clinically important because it separates a simple exposure history from the biological burden of illness. A past infection by itself did not map onto poorer test performance, but more severe symptom courses did. In practical terms, that means the degree of acute illness may matter more than the mere fact of having been infected when it comes to later cognitive testing in teenagers.
Reasoning and numeracy fell as symptom severity rose
Investigators used multivariate regression and machine learning analyses to examine self-reported COVID-19 history alongside cognitive and psychomotor performance. The cohort’s mean age was 13.5 years, placing the findings squarely in a developmental window where school performance, executive functioning, and language acquisition are still under active construction.
Across the analytical models, more severe symptom courses were significantly associated with lower logical reasoning and numerical ability. No significant gender differences appeared for those two domains. The effect was not presented as a vague statistical curiosity; the estimated performance deficit corresponded to roughly 3 to 7 points on standardized cognitive scales between the least and most severe symptom courses.
A 3- to 7-point difference may look modest on paper, but in an adolescent classroom it can decide whether a student keeps pace with new material or quietly falls behind. For reasoning and numeracy, even a small decrement can affect test scores, confidence, and the ability to absorb cumulative instruction.
Verbal ability showed a sex-specific signal
The most striking domain-specific finding involved verbal ability. Lower verbal performance emerged specifically among girls, while the reasoning and numerical findings did not show significant gender differences.
This sex-specific association does not prove a distinct biological mechanism, but it does raise the possibility that post-infectious effects may intersect with developmental timing, neuroimmune responses, or the demands of language-based tasks in different ways across sexes. In a school setting, verbal weakness can be harder to notice than a clear motor problem, yet it can affect classroom participation, reading comprehension, written expression, and retrieval of learned information.
Machine learning added nuance beyond linear regression
The machine learning models largely reinforced the regression findings for reasoning and verbal ability. They also identified fine finger dexterity and long-term memory as showing non-linear associations with symptom severity relative to the linear regression coefficients.
That non-linearity matters because not every neurocognitive effect follows a straight dose-response curve. Some domains may remain relatively stable across mild illness and then change more abruptly once symptom burden crosses a threshold. Fine finger dexterity and long-term memory belong to different functional systems than reasoning or verbal ability, yet both can influence everyday academic performance, from handwriting and device use to retention of newly learned material.
The study does not claim that every adolescent with prior COVID-19 will show measurable cognitive change. It does show that symptom burden lined up with lower scores in specific domains, and that the relationship was not limited to one test format or one analytic strategy.
Why a respiratory virus might affect neurodevelopment
SARS-CoV-2 can affect cognition in adulthood, and adolescence may be especially vulnerable because the brain is still maturing. Neurodevelopment during this period relies on ongoing synaptic refinement, network efficiency, and integration across frontal, temporal, and parietal systems. An acute systemic illness can plausibly disrupt those processes through inflammatory signaling, sleep disturbance, fatigue, reduced school attendance, or a combination of all four.
The abstract does not measure inflammatory markers, neuroimaging, or mechanism-specific biomarkers, so any biological explanation remains inferential. Even so, the clinical logic is straightforward: if symptom severity reflects greater systemic disruption, then more severe illness could reasonably leave a larger imprint on domains that depend on attention, working memory, and efficient information processing.
What the data can and cannot say
This was a self-reported COVID-19 history analysis, so infection occurrence and symptom severity were not established by virologic testing in the abstract. Self-report introduces uncertainty, and cross-sectional performance testing cannot prove that the infection caused the observed differences. Pre-existing differences in health, schooling, sleep, socioeconomic context, or learning environment could also shape cognitive scores.
Even with those limits, the cohort is large enough to make the signal hard to ignore. The results point to a specific clinical question rather than a vague concern: whether adolescents with more symptomatic COVID-19 courses should be watched more closely for changes in reasoning, verbal skills, memory, and fine motor performance.
A neurocognitive follow-up question now sits at the center
The study’s most consequential implication is not that COVID-19 always harms adolescent cognition. It is that adolescents’ COVID-19 history, especially when symptom burden was greater, was associated with measurable differences in performance during a critical developmental window.
That places neurocognitive follow-up in the same conversation as symptom resolution. If future longitudinal work confirms these associations, clinicians and educators may need to think beyond acute recovery and ask whether certain students need targeted assessment, academic support, or longer observation after more severe infection.
The unanswered issue is not whether the finding is statistically interesting. It is whether these performance differences persist, improve, or widen over time as adolescents continue through a period of rapid brain development. This cohort supplies an early warning, not a final answer.
- • Journal DOI 10.1038/s41598-026-80578-y: https://doi.org/10.1038/s41598-026-80578-y