Motor performance before, during and after COVID-19 and the role of socioeconomic background$a 10-year cohort study of 68,996 third grade children

Abstract

Background In response to the COVID-19 pandemic, various measures—including restrictions on children’s physical activities, such as national lockdowns (LD)—were implemented to contain its spread. These measures may have compromised motor development, particularly among children from lower socioeconomic backgrounds (SEBs), who are typically less active than peers from higher SEBs. This study examined the impact of COVID-19 restrictions on motor development in relation to SEB. Methods Data from 68,996 children in Germany (Age: 8.83 ± 0.56 years, range: 6.4–13.0; 35,270 female, 51.1%) assessed between 2011/2012 and 2022/2023 were analyzed from the longitudinal study ‘Berlin hat Talent’. Assessments before and after the pandemic used the German Motor Fitness Test, covering endurance, strength, coordination, and flexibility. Demographic data were collected via questionnaires; SEB was derived from official school-type classifications. Linear mixed-effect models accounted for hierarchical data: test values (level 1), motor domains (2a), participants (2b), and schools (3b). Motor performance was expressed as z-scores based on German reference percentiles. Effects of Time (pre, post LD I, post LD II), Motor Domain, and SEB (continuous, -2 to 2) were estimated, controlling for Age , Gender, and Secular Trends. Results The effect of Time was significant (p = .014, η2 < .01), with motor performance lower after LD II than prepandemic. Time × Motor Domain interaction showed motor domain-specific changes (p = .001, η2 < .01): endurance improved, while strength, coordination, and flexibility declined. Time × Motor Domain × SEB interaction was also significant (p < .001, η2 = .01), indicating that the effect of Time differed across motor domains depending on SEB. Adjusting for Secular Trends revealed that the pandemic’s overall impact (~ –4% across domains) was even stronger (p < .001, η2 = .29), with domain-specific changes to –15.47% to + 7.56%. The SEB gap slightly closed, as higher SEB groups declined more strongly (p < .001, η2 = .10).