Background: Mental health was only modestly affected in adults during the early months of the COVID-19 pandemic on the group level, but interpersonal variation was large.
Aims: We aim to investigate potential predictors of the differences in changes in mental health.
Method: Data were aggregated from three Dutch ongoing prospective cohorts with similar methodology for data collection. We included participants with pre-pandemic data gathered during 2006-2016, and who completed online questionnaires at least once during lockdown in The Netherlands between 1 April and 15 May 2020. Sociodemographic, clinical (number of mental health disorders and personality factors) and COVID-19-related variables were analysed as predictors of relative changes in four mental health outcomes (depressive symptoms, anxiety and worry symptoms, and loneliness), using multivariate linear regression analyses.
Results: We included 1517 participants with (n = 1181) and without (n = 336) mental health disorders. Mean age was 56.1 years (s.d. 13.2), and 64.3% were women. Higher neuroticism predicted increases in all four mental health outcomes, especially for worry (β = 0.172, P = 0.003). Living alone and female gender predicted increases in depressive symptoms and loneliness (β = 0.05-0.08), whereas quarantine and strict adherence with COVID-19 restrictions predicted increases in anxiety and worry symptoms (β = 0.07-0.11).Teleworking predicted a decrease in anxiety symptoms (β = -0.07) and higher age predicted a decrease in anxiety (β = -0.08) and worry symptoms (β = -0.10).
Conclusions: Our study showed neuroticism as a robust predictor of adverse changes in mental health, and identified additional sociodemographic and COVID-19-related predictors that explain longitudinal variability in mental health during the COVID-19 pandemic.
Keywords: Anxiety disorders; COVID-19; depressive disorders; epidemiology; neuroticism.