The COVID-19 pandemic represents a non-normative societal disruption that may have altered normative developmental processes in adolescence. Self-esteem, a key indicator of psychological well-being and resilience, typically undergoes changes during the transition to adolescence, yet little is known about how its developmental trajectories may have been reshaped under pandemic conditions. This study examined heterogeneity in self-esteem trajectories from late childhood to early adolescence and compared patterns between a pandemic-era cohort and a pre-pandemic comparison cohort. Additionally, we investigated early and concurrent predictors of trajectory group membership within the pandemic-era cohort.
Methods
This study used the data from two Korean longitudinal panel studies. The pandemic-era cohort (n=1,092) included children born in 2008, followed from ages 7–14 (2015–2022; not assessed in 2020), spanning pre- and mid-pandemic periods. The comparison cohort (n=1,934) included children born in 2003–2004, followed from ages 9–13 (2012–2016). Self-esteem was measured using the Rosenberg Self-Esteem Scale. Growth mixture modeling (GMM) was used to identify latent trajectory classes, and model fit supported a four-class solution in both cohorts. Within the pandemic cohort, multinomial logistic regression with multiple imputation examined predictors of class membership across early childhood, middle childhood, and early adolescence.
Results
Four trajectory classes were identified in both cohorts. According to the mean level changes, both cohorts showed a normative decline in self-esteem beginning around ages 11–12; however, differences were found in GMM classification. In the pandemic-era cohort, the classes were Low-Stable (28.2%), Persistent-Decreasing (16.8%), Late-Decreasing (50.2%), and High-Stable (4.8%). In the comparison cohort, trajectories included Low-Drop (19.4%), Moderate-Decreasing (41.2%), High-Decreasing (17.4%), and High-Stable (21.9%).
The proportion maintaining high and stable self-esteem was substantially lower in the pandemic cohort (4.8%) compared to the comparison cohort (21.9%). Conversely, a majority of the pandemic cohort (50.2%) followed a Late-Decreasing trajectory, characterized by stable self-esteem prior to age 11 followed by a sharp decline during early adolescence from age 13 (2021).
Predictor analyses revealed that lower peer attachment and more negative body image during middle childhood and early adolescence significantly increased the likelihood of belonging to vulnerable trajectories (e.g., Persistent-Decreasing: peer attachment OR=0.49–0.61; body image OR=0.45–0.62). Positive body image also predicted membership in the High-Stable group (OR=2.62 in middle childhood; OR=6.23 in early adolescence). Paternal controlling parenting in early childhood was associated with higher odds of High-Stable membership (OR=1.43), while no significant association was found for maternal parenting behaviors.
Conclusions and Implications
Findings indicate that while declines in self-esteem during early adolescence reflect a normative developmental process, the COVID-19 pandemic coincided with increased heterogeneity and a reduced likelihood of maintaining high self-esteem. The sharp divergence observed in the pandemic cohort suggests that non-normative disruptions may amplify vulnerability among adolescents, particularly by destabilizing previously adaptive trajectories.
The results underscore the need for both universal and targeted interventions, as youth who appeared well-adjusted prior to the pandemic may still experience substantial declines under changing environmental conditions, highlighting the importance of early identification and continuous developmental monitoring.
![[ Visit Client Website ]](images/banner.gif)