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.
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.
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.
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.