Changes in sleep quality and sleep disturbances in the general population from before to during the COVID-19 lockdown: A systematic review and meta-analysis

Abstract

Introduction: This systematic review and meta-analysis aims to explore changes in sleep quality and sleep disturbances in the general population from before to during the COVID-19 lockdown.

Methods: The protocol was registered in PROSPERO (CRD42021256378) and the PRISMA guidelines were followed. The major databases and gray literature were systematically searched from inception to 28/05/2021 to identify observational studies evaluating sleep changes in the general population during the lockdown with respect to the pre-lockdown period. A random effects meta-analysis was undertaken for studies reporting (a) the means of the Pittsburgh Sleep Quality Index (PSQI) global scores or the means of the sleep onset latency (SOL) times (minutes - min) before and during the lockdown, (b) the percentages of poor sleep quality before and during the lockdown, or (c) the percentages of changes in sleep quality. Subgroup analysis by risk of bias and measurement tool utilized was carried out. A narrative synthesis on sleep efficiency, sleep disturbances, insomnia and sleep medication consumption was also performed.

Results: Sixty-three studies were included. A decline in sleep quality, reflected in a pooled increase in the PSQI global scores (standardized mean difference (SMD) = 0.26; 95% CI 0.17-0.34) and in SOL (SMD = 0.38 min; 95% CI 0.30-0.45) were found. The percentage of individuals with poor sleep quality increased during the lockdown (pooled relative risk 1.4; 95% CI 1.24-1.61). Moreover, 57.3% (95% CI 50.01-61.55) of the individuals reported a change in sleep quality; in 37.3% (95% CI 34.27-40.39) of these, it was a worsening. The studies included in the systematic review reported a decrease in sleep efficiency and an increase in sleep disturbances, insomnia, and in sleep medication consumption.

Discussion: Timely interventions are warranted in view of the decline in sleep quality and the increase in sleep disturbances uncovered and their potentially negative impact on health. Further research and in particular longitudinal studies using validated instruments examining the long-term impact of the lockdown on sleep variables is needed.

Systematic review registration: https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42021256378, identifier CRD42021256378.

Keywords: COVID-19 lockdown; changes; general population; insomnia; sleep disturbances; sleep efficiency; sleep onset latency; sleep quality.

FIGURE 1

PRISMA flow diagram showing the process of study selection.

FIGURE 2

Forest plot showing pooled changes in sleep quality PSQI global score from before to during the COVID-19 lockdown. Cellini N, 2020a: Student; Cellini N, 2020b: Worker; Cellini N, 2021a: Belgian Students, Female; Cellini N, 2021b: Belgian Students, Male; Cellini N, 2021c: Belgian Regular workers, Female; Cellini N, 2021d: Belgian Regular workers, Male; Cellini N, 2021e: Belgian Remote workers, Female; Cellini N, 2021f: Belgian Remote workers, Male; Cellini N, 2021g: Belgian Unemployed/retired, Female; Cellini N, 2021h: Belgian Unemployed/retired, Male; Cellini N, 2021i: Italian Students, Female; Cellini N, 2021j: Italian Students, Male; Cellini N, 2021k: Italian Regular workers, Female; Cellini N, 2021l: Italian Regular workers, Male; Cellini N, 2021m: Italian Remote workers, Female; Cellini N, 2021n: Italian Remote workers, Male; Cellini N, 2021o:Italian Stop working, Female; Cellini N, 2021p: Italian Stop working, Male; Cellini N, 2021q: Italian Unemployed/retired, Female; Cellini N, 2021r: Italian Unemployed/retired, Male; Sella E, 2021a: Older; Sella E, 2021b: Young. Error bars = 95% confidence interval; Square boxes = individual study point estimates; and Diamond box = Pooled point estimates.

FIGURE 3

Forest plot showing pooled changes in poor sleep quality from before to during the COVID-19 lockdown. Error bars = 95% confidence interval; Square boxes = individual point estimates; and Diamond box = pooled point estimates.

FIGURE 4

Forest plot showing pooled percentages changes in sleep quality from before to during the COVID-19 lockdown: better (A), worse (B). Error bars = 95% confidence interval; Square boxes = individual point estimates; and Diamond box = pooled point estimates.

FIGURE 5

Forest plot showing pooled changes in sleep onset latency from before to during the COVID-19 lockdown. Cellini N, 2021a: Belgian Students, Female; Cellini N, 2021b: Belgian Students, Male; Cellini N, 2021c: Belgian Regular workers, Female; Cellini N, 2021d: Belgian Regular workers, Male; Cellini N, 2021e: Belgian Remote workers, Female; Cellini N, 2021f: Belgian Remote workers, Male; Cellini N, 2021g: Belgian Unemployed/retired, Female; Cellini N, 2021h: Belgian Unemployed/retired, Male; Cellini N, 2021i: Italian Students, Female; Cellini N, 2021j: Italian Students, Male; Cellini N, 2021k: Italian Regular workers, Female; Cellini N, 2021l: Italian Regular workers, Male; Cellini N, 2021m: Italian Remote workers, Female; Cellini N, 2021n: Italian Remote workers, Male; Cellini N, 2021o:Italian Stop working, Female; Cellini N, 2021p: Italian Stop working, Male; Cellini N, 2021q: Italian Unemployed/retired, Female; Cellini N, 2021r: Italian Unemployed/retired, Male. Error bars = 95% confidence interval; Square boxes = individual study effects; and Diamond box = Pooled effects.