COVID-19 abruptly raised developmental delays in preschool children across all education groups

Developmental delays among preschool children in one German region rose abruptly in the first cohort assessed after the COVID-19 pandemic began, and the increase affected children of less-educated and better-educated parents alike, according to a ten-year analysis of mandatory school entry examinations published in Archives of Public Health. The study, led by Timo-Kolja Pförtner and colleagues, covered ten consecutive annual cohorts from 2015 to 2024 in the Aachen region of North Rhine-Westphalia and focused on children whom kindergartens or schools had identified as being at risk of developmental delays before their examination. Across the decade, children whose parents had lower educational attainment consistently showed more delays than their peers, yet the pandemic-era disruption did not widen that gap in absolute terms. Instead, the sudden rise in delays was proportionally larger among children with moderate and high parental education, which narrowed inequality on a relative scale while leaving the absolute difference between groups essentially unchanged.

The evidence comes from the School Entry Examination, a compulsory assessment carried out by the local health authority in Germany to evaluate a child’s physical, mental, and health readiness for school. In Aachen, the examination typically takes place two to twelve months before school entry and includes demographic data, vaccination status, medical history, vision and hearing tests, developmental assessment, a physical examination, and a parental counseling interview. The analysis restricted itself to children reported as at risk of developmental delays, a group that made up roughly 32 percent of the roughly 4,800-child annual enrollment cohorts, totaling 15,299 children between 2015 and 2024. This restriction was necessary because pandemic-related regulations led the Aachen public health department to examine only this at-risk group in the 2021 school entry cohort, so limiting all years to comparable children preserved consistency across the decade.

Developmental delays were assessed across six domains considered central to coping with school requirements: physical coordination, selective attention, visual-motor skills, visual perception, knowledge of numbers and quantities, and speech and language. The school physician’s judgment drew on a standardized and validated screening tool known as SOPESS, the sociopediatric developmental screening for school entry examinations, together with a qualitative assessment incorporating previous findings. For each domain, the researchers coded a binary indicator: one for a delayed ability or diagnosed disability, regardless of whether treatment had begun, and zero for an inconspicuous finding. The primary outcome was the count of delayed domains, ranging from zero to six. On average, children in the analytic sample showed 1.44 delays, with a standard deviation of 1.58, and this information was missing for 14.9 percent of the sample.

Parental education was measured from information provided by legal guardians on their schooling and vocational training, converted to an eight-point index and grouped into low, moderate, and high categories, with the highest level in the household used for the analysis. Because the analytic sample was selected rather than the full cohort, the researchers applied stabilized inverse probability weighting to address potential selection bias, modeling the probability of being reported at risk on characteristics available for the entire cohort in all years. Missing values were handled through multiple imputation by chained equations carried out on the full school entry cohort of 48,121 children, with thirty imputed datasets pooled using Rubin’s rules. The weights ranged from 0.36 to 3.79 with a mean of 1.01, and after weighting, the characteristics in the model matched the full cohort to within half a percentage point.

To separate an abrupt shift from a gradual acceleration, the team used a segmented, interrupted time series Poisson regression with robust standard errors, adjusted for the child’s age at examination, sex, primary language spoken at home in the first four years of life, nationality, family structure, and kindergarten attendance. The model included a linear trend across the pre-pandemic cohorts, a level shift beginning with the 2021 cohort, a change in slope from that cohort onward, and interactions of each term with parental education. Inference on the time structure rested on joint hypothesis tests rather than individual coefficients. The break was placed between the 2020 cohort, the last examined before the pandemic, and the 2021 cohort, the first examined after its onset, because no examinations took place in the Aachen region between March and August 2020.

The results showed a stable and substantial educational gradient throughout the observation period. At the child-weighted mean cohort of the pre-pandemic period, children with low parental education were expected to show 1.766 delays, compared with 1.257 for moderate and 0.879 for high parental education, corresponding to significant differences of 0.510 and 0.887 delays. Before the pandemic, delays increased significantly by 0.057 per cohort among children with low parental education and by 0.046 among those with moderate parental education, while the increase of 0.033 in the high-education group was not statistically significant. Expressed relatively, these pre-pandemic trends ran in parallel, rising by about three to four percent per cohort in all educational groups, and the statistical tests found no divergence in trajectories before the pandemic.

That steady climb was interrupted with the 2021 cohort, when delays shifted upward sharply in all three educational groups before declining again in subsequent years without returning to pre-pandemic levels. The absolute shift in 2021 amounted to 0.416 delays among children with low parental education, 0.689 among those with moderate parental education, and 0.324 among those with high parental education, and each of these shifts was statistically significant. In the 2021 cohort, the mean number of delays reached 1.66, the highest observed over the entire period. After the break, delays declined significantly by 0.093 per cohort among children with low parental education and by 0.148 among those with moderate parental education, while the decline of 0.031 in the high-education group was not significant. Notably, among children with high parental education, delays remained stable across the observed period apart from the significant 2021 shift itself.

The pandemic period produced a paradoxical pattern in how inequality looked depending on the yardstick. Relative inequality, expressed as the ratio of expected delays between educational groups, narrowed significantly between the two periods: the ratio for moderate versus low parental education rose from 0.711 to 0.786 and for high versus low from 0.498 to 0.582, changes of 1.105 and 1.169 that were jointly significant. This narrowing occurred because delays increased proportionally more among children with moderate and high parental education, whose relative rise was 1.294 and 1.369, than among children with low parental education, whose rise was 1.171. On the absolute scale, however, the increases of 0.302, 0.369, and 0.324 delays for the low, moderate, and high education groups were of comparable size, so the absolute gap remained largely unchanged at 0.442 and 0.865 delays, with a nonsignificant change. The authors caution that this relative narrowing may be a mathematical consequence of differing baseline levels rather than a substantive reduction in disadvantage, since a similar absolute increase amounts to a smaller proportional rise in a group that starts from a higher baseline.

Sensitivity analyses tested the robustness of these findings from several angles. The main conclusions held when complete cases were used instead of imputation, when the weights were omitted, and across five alternative definitions of the at-risk sample, including varying the cut-off date used to identify at-risk children and excluding the 2021 cohort entirely, although the moderate versus low contrast fell just short of significance in two of these analyses. When the researchers allowed a nonlinear pandemic trajectory after 2021, the reduction in relative inequality was attenuated and no longer statistically significant, indicating that this particular finding is sensitive to the functional form assumed for the post-pandemic period. Analyzed domain by domain, the reduction in relative inequality was confined to speech and language and visual-motor skills and did not survive correction for multiple testing, while no domain showed a significant change on the absolute scale.

The authors are careful about what the study can and cannot establish. Because the design used school entry cohorts over time without a separate control group, the effects of the pandemic cannot be separated from other concurrent changes, so the observed 2021 shift reflects everything that changed during that period, not only the mitigation measures. The analysis was also limited to children identified as at greater risk before the examination, so the estimates describe how delays and their social gradient developed among children brought to attention, not how frequent delays were in the cohort as a whole, and the identification relied on staff observations rather than systematic screening. The 2021 cohort was examined under a modified selection rule and at a mean age about two months higher than surrounding cohorts, which the models accounted for with a dummy variable, but altered examination strategies and reduced attendance during closures may still have constrained professionals’ observation opportunities. The data derive from a single region, parental education was the only available indicator of social position, and socioemotional development and household income were not recorded. Within those limits, the study’s central message stands: the educational gradient in developmental delays proved remarkably persistent through a global shock, the dominant feature of the period was an abrupt, population-wide disruption followed by a partial recovery that left levels elevated, and addressing deep-rooted socioeconomic disparities, the authors conclude, requires sustained structural interventions rather than temporary crisis management.

Subject of Research: Time trends in developmental delays among preschool children by parental education before and after the onset of the COVID-19 pandemic

Article Title: Has the COVID-19 pandemic altered social inequalities in child development? A time trend analysis of developmental delays in preschool children by parental education

Article References: Pförtner, T.-K., Hennemann, T., Gube, M., & Elgar, F. J. (2026). Has the COVID-19 pandemic altered social inequalities in child development? A time trend analysis of developmental delays in preschool children by parental education. Archives of Public Health, 84(1), Article 223. https://doi.org/10.1186/s13690-026-02111-x

Image Credits: AI Generated

DOI: 10.1186/s13690-026-02111-x

Keywords: COVID-19, child development, developmental delays, parental education, health inequalities, school entry examination, preschool children, interrupted time series, Germany, socioeconomic status, public health, pandemic