A systematic review and meta-analysis of longitudinal cohort studies comparing mental health before versus during the COVID-19 pandemic in 2020

Abstract

Background: Increases in mental health problems have been observed during the COVID-19 pandemic. The objectives were to examine the extent to which mental health symptoms changed during the pandemic in 2020, whether changes were persistent or short lived, and if changes were symptom specific.

Methods: Systematic review and meta-analysis of longitudinal cohort studies examining changes in mental health among the same group of participants before vs. during the pandemic in 2020.

Results: Sixty-five studies were included. Compared to pre-pandemic outbreak, there was an overall increase in mental health symptoms observed during March-April 2020 (SMC = .102 [95% CI: .026 to .192]) that significantly declined over time and became non-significant (May-July SMC = .067 [95% CI: -.022 to .157]. Compared to measures of anxiety (SMC = 0.13, p = 0.02) and general mental health (SMC = -.03, p = 0.65), increases in depression and mood disorder symptoms tended to be larger and remained significantly elevated in May-July [0.20, 95% CI: .099 to .302]. In primary analyses increases were most pronounced among samples with physical health conditions and there was no evidence of any change in symptoms among samples with a pre-existing mental health condition.

Limitations: There was a high degree of unexplained heterogeneity observed (I²s > 90%), indicating that change in mental health was highly variable across samples.

Conclusions: There was a small increase in mental health symptoms soon after the outbreak of the COVID-19 pandemic that decreased and was comparable to pre-pandemic levels by mid-2020 among most population sub-groups and symptom types.

Keywords: Anxiety; COVID-19; Coronavirus; Depression; Longitudinal; Mental health.

Fig. 1.

Study search and inclusion flow chart ‘Not a research report’ indicates that a manuscript was ineligible as it did not report on an empirical study. Excluded as unable to include in meta-analysis (N = 1) was due to authors not providing the required missing statistical information to compute effect size after having been contacted.

Fig. 2.

Forest plot of effect sizes from studies of depression, general mental health and anxiety symptoms with lower risk of bias. Analyses of more homogenous collection of studies (depression, anxiety, general mental health measures only) that were of lower risk of bias (used representative sampling, did not report inconsistent mode of survey delivery, reported no conflicts of interest) of ∼27,736 participants, SMC = .100 [95% CI: .033 to .166], z = 2.95, p = .003, I² = 98.0%). General (general population sample), MH (sample with pre-existing mental health condition).

Fig. 3.

Forest plot of effect sizes for exceeding questionnaire cut offs. General MH (measure of general mental health functioning). Marginal Odds Ratio = 1.31 [95% CI: 1.10 to 1.55], z = 3.18, p = .001, I² = 93.2%).