Psychiatric and neuropsychiatric sequelae of COVID-19 within 2 years: a multinational cohort study.

Background: The long-term psychiatric and neuropsychiatric sequelae of COVID-19 across diverse populations remain not fully understood. This cohort study aims to investigate the short-, medium-, and long-term risks of psychiatric and neuropsychiatric disorders following COVID-19 infection in five countries.

Methods: This population-based multinational network study used electronic medical records from France, Italy, Germany, and the UK and claims data from the USA. The initial target and comparator cohorts were identified using an exact matching approach based on age and sex. Individuals diagnosed with COVID-19 or those with a positive SARS-CoV-2 screening test between December 1, 2019, and December 1, 2020, were included as targets. Up to ten comparators without COVID-19 for each target were selected using the propensity score matching approach. All individuals were followed from the index date until the end of continuous enrolment or the last healthcare encounter. Cox proportional hazard regression models were fitted to estimate the risk of incident diagnosis of depression, anxiety disorders, alcohol misuse or dependence, substance misuse or dependence, bipolar disorders, psychoses, personality disorders, self-harm and suicide, sleep disorders, dementia, and neurodevelopmental disorders within the first 6 months (short-term), 6 months to 1 year (medium-term), and 1 to 2 years (long-term) post-infection.

Results: A total of 303,251 individuals with COVID-19 and 22,108,925 individuals without COVID-19 from five countries were originally included. Within the first 6 months, individuals with COVID-19 had a significantly higher risk of any studied disorders in all databases, with Hazard Ratios (HRs) ranging from 1.14 (95% CI, 1.07-1.22) in Germany to 1.89 (1.64-2.17) in Italy. Increased risks were consistently observed for depression, anxiety disorders, and sleep disorders across almost all countries. During the medium- and long-term periods, higher risks were observed only for depression (medium-term: 1.29, 1.18-1.41; long-term: 1.36, 1.25-1.47), anxiety disorders (medium-term: 1.29, 1.20-1.38; long-term: 1.37, 1.29-1.47), and sleep disorders (medium-term: 1.10, 1.01-1.21; long-term: 1.14, 1.05-1.24) in France, and dementia (medium-term: 1.65, 1.28-2.10) in the UK.

Conclusions: Our study suggests that increased risks of psychiatric and neuropsychiatric outcomes were consistently observed only within, and not after, the 6-month observation period across all databases, except for certain conditions in specific countries.