Mental illness after bereavement before and during the COVID-19 pandemic in Sweden: A matched cohort study

Bereavement is associated with an increased risk of mental illness. The COVID-19 pandemic caused excess mortality, and may have exacerbated the mental health impact of bereavement due to social restrictions and reduced healthcare access. Using Swedish national register data, this study aimed to compare the risk of mental illness following bereavement before (2018–2019) and during (2020–2021) the pandemic, exploring how the pandemic might have modified the psychological impact of bereavement and identifying high-risk groups.We conducted a nationwide matched cohort study including (1) 3,840,845 individuals (349,168 bereaved) before the pandemic, and (2) 5,132,988 individuals (466,636 bereaved) during the pandemic. Mental illness was defined as the first occurrence of any psychiatric diagnosis or suicidal behavior during each period. Multivariable Cox regression was used to estimate hazard ratios (HRs) with 95% confidence intervals (CIs). We found that bereaved individuals had a significantly higher risk of mental illness compared to non-bereaved individuals in both periods (before pandemic: HR 1.42, 95%CI 1.34-1.49; during pandemic: HR 1.34, 95%CI 1.28-1.39). Bereaved individuals younger than 30 years had markedly higher risks of psychiatric disorders in the pre-pandemic period compared to the pandemic period. Higher risks of incident psychiatric disorders were observed for loss of a child or spouse, compared to loss of a sibling or parent, as well as for loss due to accident or suicide as compared to other causes. Furthermore, bereavement due to COVID-19 was associated with an increased risk of mental illness during the pandemic period (HR 1.38, 95% CI 1.21-1.59). In conclusion, bereavement was consistently associated with an increased risk of mental illness, before and during COVID-19 pandemic, although young individuals (<30 years) seemed more affected before the pandemic. However, further research in settings with a different pandemic burden and/or mitigation strategies is needed to assess the generalizability of our findings beyond Sweden.