This prospective analysis included 1,124 children from low-income, racially diverse families followed across three U.S. cities: New York City, Pittsburgh, and Tulsa. At the time the pandemic began (March 2020), children were aged approximately 3–7 years. The sample was balanced by sex (50% female) and comprised 48% Hispanic/Latine, 30% Black, and 22% White/Other participants. The study intentionally focused on populations disproportionately affected by pandemic-related hardships due to low income and systemic racism.
Researchers used repeated behavioral assessments conducted between 2016 and 2023, providing prospective pre- and post-pandemic data for the same children. The paper reports changes in multiple behavioral domains across this multi-year window. Figures in the full text illustrate trajectories before and after COVID-19 onset for each city, with marked time points for March 2020 and subsequent years.
The study examined domain-specific changes in children’s functioning over time. Outcomes discussed in the abstract and plain-language summary include aggression, emotional reactivity, attention problems, anxiety, withdrawn symptoms, prosocial behavior, and executive functioning (concentration/attention skills). The authors compare intercepts and slopes of behavioral trajectories before versus after pandemic onset to identify both level shifts and changes in growth rates. Specific statistical models and effect sizes were not provided in the report excerpt.
In New York, which had a sample characterized as high in Hispanic representation, the aggregate pattern was stability or slight improvement following pandemic onset. Levels and growth rates for aggression and emotional reactivity after March 2020 resembled pre-pandemic trajectories. Measures of attention problems, anxiety, and withdrawn symptoms declined in level after the pandemic began. However, the rate of growth for anxiety increased after the pandemic in New York, described as a catch-up toward pre-pandemic levels over time. Figure annotation in the full report indicates statistically significant changes in intercepts and slopes at specified post-pandemic time points.
In Pittsburgh, where the sample was reported as largely Black, the observed pattern differed: the measured behaviors rose following pandemic onset. The increase after March 2020 affected multiple domains included in the study. Importantly, most of these elevations in problem behaviors reportedly returned to expected or pre-pandemic levels within approximately two years, indicating a degree of recovery by the later follow-up points. The full report’s figures show timing of statistically significant intercept and slope changes specific to Pittsburgh’s trajectory.
In Tulsa, results showed declines in prosocial behavior and a slowing of its growth rate after the pandemic began. A pre-pandemic declining trend in disruptive behaviors slowed following COVID-19 onset, and aggregate executive functioning scores fell in level. By contrast, concentration and attention-related measures in Tulsa remained largely steady across the study period. Figure legends identify specific semesters and seasons (e.g., Fall, Spring) where intercept or slope changes reached statistical significance.
Across the three cities, the authors report heterogeneous outcomes rather than a uniform worsening of young children's behavior after the onset of COVID-19. New York showed stability or modest improvement on several indices, Pittsburgh exhibited initial worsening followed by partial recovery within two years, and Tulsa had notable declines in prosocial and executive functioning measures with relative stability in attention. The paper emphasizes that regional variation—potentially reflecting local pandemic conditions, policy responses, community resources, or demographic composition—should be considered when assessing pandemic impacts on child mental health and behavior.
Key strengths reported include the prospective design with repeated measures spanning pre- and post-pandemic years and the focus on low-income, racially diverse urban samples that were disproportionately impacted by the pandemic. The authors note the importance of considering geographic heterogeneity in pandemic effects. The provided abstract and summary do not include detailed descriptions of statistical methods, recruitment procedures, potential mediators/moderators, or effect sizes; those methodological specifics and additional contextual explanations are available in the full text. The authors call for future research to investigate causes of the observed regional variation.