Despite the multitude of clinical manifestations of post-acute sequelae of SARS-CoV-2 infection (PASC), studies applying statistical methods to directly investigate patterns of symptom co-occurrence and their biological correlates are scarce.

We assessed 30 symptoms pertaining to different organ systems in 749 adults (age = 55 ± 14 years; 47% female) during in-person visits conducted at 6–11 months after hospitalization due to coronavirus disease 2019 (COVID-19), including six psychiatric and cognitive manifestations. Symptom co-occurrence was initially investigated using exploratory factor analysis (EFA), and latent variable modeling was then conducted using Item Response Theory (IRT). We investigated associations of latent variable severity with objective indices of persistent physical disability, pulmonary and kidney dysfunction, and C-reactive protein and D-dimer blood levels, measured at the same follow-up assessment.

The EFA extracted one factor, explaining 64.8% of variance; loadings were positive for all symptoms, and above 0.35 for 16 of them. The latent trait generated using IRT placed fatigue, psychiatric, and cognitive manifestations as the most discriminative symptoms (coefficients > 1.5, p < 0.001). Latent trait severity was associated with decreased body weight and poorer physical performance (coefficients > 0.240; p ⩽ 0.003), and elevated blood levels of C-reactive protein (coefficient = 0.378; 95% CI 0.215–0.541; p < 0.001) and D-dimer (coefficient = 0.412; 95% CI 0.123–0.702; p = 0.005). Results were similar after excluding subjects with pro-inflammatory comorbidities.

Different symptoms that persist for several months after moderate or severe COVID-19 may unite within one latent trait of PASC. This trait is dominated by fatigue and psychiatric symptoms, and is associated with objective signs of physical disability and persistent systemic inflammation.

There is mixed evidence on increasing rates of psychiatric disorders and symptoms during the coronavirus disease 2019 (COVID-19) pandemic in 2020. We evaluated pandemic-related psychopathology and psychiatry diagnoses and their determinants in the Brazilian Longitudinal Study of Health (ELSA-Brasil) São Paulo Research Center.

Between pre-pandemic ELSA-Brasil assessments in 2008–2010 (wave-1), 2012–2014 (wave-2), 2016–2018 (wave-3) and three pandemic assessments in 2020 (COVID-19 waves in May–July, July–September, and October–December), rates of common psychiatric symptoms, and depressive, anxiety, and common mental disorders (CMDs) were compared using the Clinical Interview Scheduled-Revised (CIS-R) and the Depression Anxiety Stress Scale-21 (DASS-21). Multivariable generalized linear models, adjusted by age, gender, educational level, and ethnicity identified variables associated with an elevated risk for mental disorders.

In 2117 participants (mean age 62.3 years, 58.2% females), rates of CMDs and depressive disorders did not significantly change over time, oscillating from 23.5% to 21.1%, and 3.3% to 2.8%, respectively; whereas rate of anxiety disorders significantly decreased (2008–2010: 13.8%; 2016–2018: 9.8%; 2020: 8%). There was a decrease along three wave-COVID assessments for depression [β = −0.37, 99.5% confidence interval (CI) −0.50 to −0.23], anxiety (β = −0.37, 99.5% CI −0.48 to −0.26), and stress (β = −0.48, 99.5% CI −0.64 to −0.33) symptoms (all ps < 0.001). Younger age, female sex, lower educational level, non-white ethnicity, and previous psychiatric disorders were associated with increased odds for psychiatric disorders, whereas self-evaluated good health and good quality of relationships with decreased risk.

No consistent evidence of pandemic-related worsening psychopathology in our cohort was found. Indeed, psychiatric symptoms slightly decreased along 2020. Risk factors representing socioeconomic disadvantages were associated with increased odds of psychiatric disorders.

Depression is highly prevalent and marked by a chronic and recurrent course. Despite being a major cause of disability worldwide, little is known regarding the determinants of its heterogeneous course. Machine learning techniques present an opportunity to develop tools to predict diagnosis and prognosis at an individual level.

We examined baseline (2008–2010) and follow-up (2012–2014) data of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), a large occupational cohort study. We implemented an elastic net regularization analysis with a 10-fold cross-validation procedure using socioeconomic and clinical factors as predictors to distinguish at follow-up: (1) depressed from non-depressed participants, (2) participants with incident depression from those who did not develop depression, and (3) participants with chronic (persistent or recurrent) depression from those without depression.

We assessed 15 105 and 13 922 participants at waves 1 and 2, respectively. The elastic net regularization model distinguished outcome levels in the test dataset with an area under the curve of 0.79 (95% CI 0.76–0.82), 0.71 (95% CI 0.66–0.77), 0.90 (95% CI 0.86–0.95) for analyses 1, 2, and 3, respectively.

Diagnosis and prognosis related to depression can be predicted at an individual subject level by integrating low-cost variables, such as demographic and clinical data. Future studies should assess longer follow-up periods and combine biological predictors, such as genetics and blood biomarkers, to build more accurate tools to predict depression course.