Protective and risk factors of anxiety in children and adolescents during COVID-19: A systematic review and three level meta-analysis

Abstract

Background: In order to gain a deepened understanding of the impact of public health emergency and to develop effective interventions and preventions, this study aimed to evaluate risk and protective factors associated with anxiety in children and adolescents and to explore potential moderators in the background of COVID-19 within the framework of socio-ecological model. Methods: A literature search was conducted in Web of Science, MEDLINE, PubMed, Scopus, EBSCO, ScienceDirect, Emerald, and CNKI for studies published from early 2020 to early 2023 that examined factors associated with anxiety among children and adolescents with age range of 6 to 17 years old. Random effects models and a three-level meta-analytic approach were used. Results: In total, 141 articles and 1,018,171 subjects were included, with 1002 effect sizes extracted. 32 protective factors and 48 risk factors were examined, yielding significance for 14 protective factors and 29 risk factors, including individual factors (e.g., gender, logOR = −0.37, 95 % CI [−0.47, −0.27], p < 0.001; age, logOR = −0.12, 95 % CI [−0.22, −0.02], p = 0.02; emotional functioning, logOR = −1.45, 95 % CI [−1.84, −1.05], p < 0.001; pre-existing condition logOR = 0.94, 95 % CI [0.58, 1.30], p < 0.001; electronic device or internet addiction, logOR = 1.81, 95 % CI [0.74, 2.88], p < 0.001), family factors (e.g., family socioeconomic status, logOR = −0.25, 95 % CI [−0.39, −0.10], p < 0.001; family functioning, logOR = −1.31, 95 % CI [−1.60, −1.02], p < 0.001; anxiety level of caregiver, logOR = 1.06, 95 % CI [0.75, 1.37], p < 0.001), community factors (e.g., overall social support, logOR = −0.93, 95 % CI [−1.84, −1.05], p < 0.001; school burden, logOR = 0.56, 95 % CI [0.21, 0.90], p = 0.002), and COVID-19-related factors (e.g., higher exposure risk in local community or city, logOR = 0.48, 95 % CI [0.17, 0.78], p = 0.002; distant learning, logOR = 0.73, 95 % CI [0.19, 1.28], p = 0.008; COVID-19-related distress, logOR = 1.42, 95 % CI [0.55, 2.29], p = 0.001;). The majority of studies showed no publication bias. Age group moderated the relationship between gender and level of anxiety (F (1,96) = 4.42, p = 0.038), and no other moderator showed significance. Limitations: This study does not reveal causality but correlation in nature, and our findings should be interpretated with caution. Conclusions: Public health emergencies could bring challenges to the mental health of children and adolescents. Prevention and intervention strategies for children and adolescents with high risks, and family-based and community-based programs should be encouraged to buffer the adverse impact on children and adolescents. This study has been prospectively registered at PROSPERO (registration number: CRD42022316746).