Background and aims
Little is known about post-COVID-19 neurological symptoms. We aimed to assess neurological manifestations of post-COVID-19 patients and their relation to severity.
Methods
We performed a cross-sectional analysis of 779 consecutive COVID-19 patients (PCR-confirmed) in the post-acute stage (585 women, median age 42, median interval post-infection 48 days). All participants filled the form (https://forms.gle/vgoyHQ6wftkucgtM8). We performed a multinomial logistic regression with cross-entropy optimization to predict subject cluster using sex, treatment modality (outpatient (696), non-ICU-inpatient (52) and ICU-inpatient (31)) and age groups as regressors. SciPy 1.5.2 and Scikit-Learn 0.23.2 Python packages were used for the analyses.
Results
Most of the participants were polysymptomatic (18.6% were asymptomatic, 19.6% had one symptom, 17.1% had two symptoms, and 44.7% presented three or more symptoms), including fatigue (48%), headache (34%) and memory problems (29.8%). Cluster analysis divided the 779 individuals into five clusters of subjects (anosmia predominant, oligosymptomatic, polysymptomatic, headache predominant and fatigue/memory predominant), and four clusters of symptoms (headache, fatigue/memory issues, anosmia/ageusia and miscellaneous symptoms). Regarding treatment, the non-ICU inpatients (compared to the outpatients) were less likely to be in the anosmia-predominant cluster (p = 0.04); the ICU inpatients (compared to the outpatients) had a higher probability of being in the fatigue-memory cluster (p = 0.001). Women presented a higher chance to be in the headache cluster (p < 0.001) and a lesser chance to be in the oligosymptomatic (p < 0.001) and fatigue-memory predominant (p = 0.02) groups.
Conclusions
Anosmia seems associated with milder manifestations, and the fatigue-memory group appears associated with ICU admission. Women presented more headache and were more polysymptomatic than men.