Long-term neurological and cognitive impact of COVID-19: a systematic review and meta-analysis in over 4 million patients

BACKGROUND: Neuropsychiatric symptoms emerged early in the COVID-19 pandemic as a key feature of the virus, with research confirming a range of neuropsychiatric manifestations linked to acute SARS-CoV-2 infection. However, the persistence of neurological symptoms in the post-acute and chronic phases remains unclear. This meta-analysis assesses the long-term neurological effects of COVID-19 in recovered patients, providing insights for mental health service planning. METHODS: A comprehensive literature search was conducted across five electronic databases: PubMed, Scopus, Web of Science, EBSCO, and CENTRAL, up to March 22, 2024. Studies evaluating the prevalence of long-term neurological symptoms in COVID-19 survivors with at least six months of follow-up were included. Pooled prevalence estimates, subgroup analyses, and meta-regression were performed, and publication bias was assessed. RESULTS: The prevalence rates for the different symptoms were as follows: fatigue 43.3% (95% CI [36.1-50.9%]), memory disorders 27.8% (95% CI [20.1-37.1%]), cognitive impairment 27.1% (95% CI [20.4-34.9%]), sleep disorders 24.4% (95% CI [18.1-32.1%]), concentration impairment 23.8% (95% CI [17.2-31.9%]), headache 20.3% (95% CI [15-26.9%]), dizziness 16% (95% CI [9.5-25.7%]), stress 15.9% (95% CI [10.2-24%]), depression 14.0% (95% CI [10.1-19.2%]), anxiety 13.2% (95% CI [9.6-17.9%]), and migraine 13% (95% CI [2.2-49.8%]). Significant heterogeneity was observed across all symptoms. Meta-regression analysis showed higher stress, fatigue, and headache in females, and increased stress and concentration impairment with higher BMI. CONCLUSIONS: Neurological symptoms are common and persistent in COVID-19 survivors. This meta-analysis highlights the significant burden these symptoms place on individuals, emphasizing the need for well-resourced multidisciplinary healthcare services to support post-COVID recovery. REGISTRATION AND PROTOCOL: This meta-analysis was registered in PROSPERO with registration number CRD42024576237.