Abstract
Even after a mild episode of coronavirus disease 2019 (COVID-19), many patients suffer from persistent symptoms that can impair their quality of life for months. The potential role of psychological mechanisms in this post-COVID-19 condition, often referred to as long COVID, has been discussed early in the pandemic. Evidence supporting this hypothesis is now accumulating. First, a history of anxiety or depression is now an established risk factor for post-COVID-19 condition with a dose-response relationship and effect size similar to those of other known risk factors. This association extends to other forms of psychological distress, including perceived stress and loneliness. Second, specific beliefs about COVID-19 have been associated with the risk of subsequent similar symptoms, occurring weeks to several months later. Other studies, which have yet to be replicated, suggest an influence of the context of the initial infection (first versus second wave of the pandemic, before the emergence of significant variants and vaccination) and the trust in various sources of information about COVID-19 on the risk of subsequent symptoms. Bayesian models of perception can account for these results particularly well within a theoretical framework similar to that advanced for functional somatic disorders, integrating increased symptom expectations with decreased perception of the body internal state (interoception) and intolerance of uncertainty in the context of symptoms initially triggered by an infectious episode. These psychological mechanisms should obviously not be considered as exclusive. However, since they are modifiable, they could be targeted in clinical trials, within an integrative and multidisciplinary approach.