# Evaluating contributions of neuropsychological, psychiatric, and inflammatory processes to the expression of cognitive symptoms in post-acute COVID-19 syndrome

## TL;DR

This study explores how brain function, mental health, and inflammation relate to cognitive symptoms in long COVID patients.

## Contribution

The study identifies neuropsychological test performance as a stronger predictor of cognitive symptoms than psychiatric or inflammatory factors in PACS.

## Key findings

- 36% of variance in self-reported cognitive symptoms was explained by anxiety, premorbid intelligence, and neuropsychological test performance.
- Observed cognitive symptoms were explained by specific neuropsychological tests, accounting for 33% of variance.
- Depression, fatigue, and inflammatory cytokines did not significantly contribute to cognitive symptom models.

## Abstract

Post-acute COVID-19 syndrome (PACS) has been widely associated with cognitive symptoms, however, the nature and severity of effects on cognitive function have been difficult to establish amid other aspects of PACS symptomatology. The current study used a regression modeling approach to parse unique and combined contributions of neuropsychological test performance, psychiatric symptoms, and inflammatory cytokine levels in predicting cognitive symptom severity, as measured by questionnaire responses from patient and observer perspectives.

Forty-one patients presenting to a university medical center neurology clinic with cognitive complaints ≥ 4 months after COVID-19 symptom onset were included in the analysis.

Although pre-morbid cognitive status was estimated to be at or above-average across participants, nearly 50% performed below expectation on three or more neuropsychological tests. Subjective Cognitive Decline Questionnaire (SCD-Q) ratings were clinically elevated, both from patient (MyCog) and observer (TheirCog) perspectives, yet bivariate relationships with neuropsychological and other measures of PACS symptomatology were non-significant. When combined in regression models, 36% of variance in MyCog score was explained by measures of anxiety, premorbid intelligence, and current neuropsychological test performance. TheirCog scores were explained by a unique set of neuropsychological tests, accounting for 33% of variance cumulatively. Measures of depression, fatigue, and inflammatory cytokines concentrations did not enter either model.

Taken together, cognitive sequelae of PACS appear to be rooted in changes in brain function that are detectable by objective neuropsychological testing. Although comorbidities associated with PACS can contribute to the experience of cognitive symptoms, we find cognitive symptoms, whether self-reported or observed, to be more directly associated with neuropsychological test performance than ongoing fatigue, psychiatric symptoms, or inflammatory processes.