ABSTRACT
Background and aims: Register studies and cohort analyses of clinical data are essential to study neurological manifestations of COVID-19 at a large scale. Methods: We analyzed neurological manifestations in COVID-19 patients, diagnosed before Aug 25th 2020, and registered in the European multinational LEOSS registry. Results: Of the 3127 COVID-19 patients, 95.2% were hospitalized. In 54.4% at least one neurological symptom, and in 3.3% a new neurological complication occurred. Preexisting neurological comorbidities were reported in 18.1% of the patients. Neurological symptoms were excessive tiredness (27.6%), headache (15.3%), nausea/emesis (14.0%), muscular weakness (13.2%), smell (6.9%), taste disorder (8.3%) and delirium (6.3%). Intracerebral bleeding occurred in 1.2%, ischemic stroke in 0.5%, and meningitis/ encephalitis in 0.4%. Overall, the death rate was 17.5%. It was higher in patients with the following neurological comorbidities: dementia 38.0%, movement disorders 32.8%, and prior cerebrovascular disease 32.3%. A multivariable logistic regression model found age (OR 1.53), cardiovascular diseases (OR 1.74), muscle weakness (OR 1.40), pulmonary diseases (1.49) and male gender (OR 1.52) to be associated with a significantly increased risk for a critical COVID-19 disease course, failed recovery, and death. Conclusion: The neurological manifestations revealed in COVID-19 patients of this study are mostly in agreement with previously published data. Several neurological conditions, such as prior cerebrovascular diseases or dementia appeared to be associated with a higher risk in unadjusted analyses, which was not confirmed in a multivariable analysis adjusting for confounding variables such as age and sex. These findings contrast previously published studies and stress the importance of considering putative confounds in medical statistics carefully.