Abstract
Objective
This cohort study examined whether documented coronavirus disease 2019 in young children affects the need for neurodevelopmental treatment and its association with developmental diagnoses and other medical conditions.
Methods
Children with and without coronavirus disease 2019 were compared before and after an index date, defined as the first confirmed severe acute respiratory syndrome coronavirus 2 infection for participants with an infection and a matched reference date for controls. Data were obtained from a large health maintenance organization.
Results
The study included 227,728 children born between 1 January 2017 and 31 January 2022 who were followed up from birth until 1 January 2024. Before the index date, there was no significant difference in the neurodevelopmental treatment between the two groups. After diagnosis, the affected children received significantly more treatments than those without an infection. Diagnosed children also exhibited higher rates of ear, nose, and throat; respiratory; and gastrointestinal conditions both before and after diagnosis. Emotional difficulties, including anxiety and acute stress reactions, were more common post-infection, whereas no increase in developmental diagnoses, such as developmental delay and autism spectrum disorder, was observed.
Conclusion
Observations on the same cohort before and after severe acute respiratory syndrome coronavirus 2 exposure suggest that the infection rather than adverse neurodevelopmental outcomes prompted greater parental concern and increased therapeutic support.
Keywords
Introduction
The long-term developmental and health outcomes in children following the coronavirus disease 2019 (COVID-19) pandemic are critical. The pandemic may have impacted this population through direct exposure, illness, and/or prolonged periods of isolation as well as its consequences. Due to various limitations, many reports have utilized pre- and post-pandemic data to assess the impact of the COVID-19 pandemic on children’s development. 1 A systematic review of neurodevelopmental impairment revealed an increase in communication disorders among young infants during the pandemic compared with that in prepandemic infants of the same age, regardless of the mothers’ exposure; however, there were no significant differences in other areas of development. 2 Compared with prepandemic data, the language development of infants born and raised during the pandemic was impaired.3–5 Infants not exposed to the virus in utero or during the first months of life reported significantly lower developmental scores than prepandemic infants, particularly in the fine motor skills and communication domains.6–9 Only a few studies have investigated the direct effect of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection on child development outcomes.10,11 Twenty infants with SARS-CoV-2 infection during the neonatal period were assessed after 21 months. Nine children (45%) demonstrated mild delay, and two (10%) showed moderate delay, mostly in expressive and receptive language domains as well as fine motor skills. 10 A similar study has reported no differences in the neurodevelopmental outcomes at 18 months among 40 neonates with SARS-CoV-2 infection compared with controls. 11
One cohort study further supports the evidence of long-term COVID-19 effects in children, reporting significant increases in referrals for mental health services and medication prescriptions among the affected children, indicating long-term morbidity. 12
This comprehensive cohort study aimed to describe the potential influence of documented COVID-19 infection in children on the need for treatment at child development centers as well as on various developmental diagnoses and other medical conditions in this population. We hypothesized that the number of treatments and developmental diagnoses would be higher at the 2-year follow-up in children with COVID-19 than in those without COVID-19.
Methods
Data
The study population comprised all children born between 1 January 2017 and 31 January 2022 within a single health maintenance organization (HMO), which provides medical services to over 2.7 million members. In Israel, all citizens must enroll in 1 of 4 competing health funds, which all provide the same basket of medical services with regulated copayments or free of charge. We extracted comprehensive deidentified information from the HMO electronic medical records (EMRs), including data on occupational therapy, speech and language pathology, physiotherapy, and psychosocial interventions (such as social work and psychological consultations). Additionally, we collected information regarding socioeconomic status (SES) and other diagnoses made by primary pediatricians, including SARS-CoV-2 infection and congenital diseases. SARS-CoV-2 infection was defined based on physician notes. The source of the diagnosis, which could be a self-reported home test or a clinic-based test, was not available.
The Ethical Committee of Maccabi Health Services approved this study.
Outcomes of interest
The primary outcomes of interest were the total number and monthly average of neurodevelopmental treatments and visits (NDTs) after the index date. NDTs encompass any neurodevelopmental intervention, including occupational therapy, speech and language pathology, physiotherapy, and psychosocial interventions (such as social work and psychological consultations). The index date refers to the date of the first confirmed SARS-CoV-2 infection in the child as documented by the primary physician. For children without an infection, the index date was matched with those of participants without an infection to ensure consistency. As the total number of NDTs depends on the length of the follow-up of each participant, we considered an additional outcome of normalized NDTs. To calculate the normalized monthly average of NDTs, the NDT value was divided by the time (in months) since the index date, allowing for comparisons for a varying follow-up period across the population. By monitoring the progression of health and developmental outcomes over time, we aimed to identify any long-term effects of SARS-CoV-2 infection, comparing the health status of children with and without infection during and after the specified period.
Secondary outcomes of interest included neurological and developmental disorders International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) 13 as well as other medical diagnoses, in both groups, both before and after the index date.
Sample matching approach
To statistically analyze the effect of SARS-CoV-2 in the population, we compared the outcomes in children with an infection and matched controls comprising individuals without an infection, using a sample matching method. 14 This approach was designed to minimize potential bias due to confounding variables, ensuring that any observed differences in outcomes could be attributed to the treatment, in this case, NDTs, rather than underlying disparities in participant characteristics. In this study, individuals with an infection (cases) were matched with those without an infection (controls) based on specific criteria, including sex, SES, date of birth (within 10 days), presence of congenital diseases (if any), and number of NDTs before the index date. We used exact matching (except for date of birth and number of NDTs) rather than propensity matching. Therefore, instead of defining a caliper, we allowed some deviation in the exact matching for date of birth and number of NDTs, as described. The index date for each case was assigned to match that of the corresponding control to maintain precise temporal alignment between the groups. Matching of neurodevelopmental treatment utilization prior to the index date was based on the total number of treatment visits, regardless of treatment type or intensity, to balance overall use of developmental services and healthcare engagement between groups.
Statistical analyses
Comprehensive statistical analyses were performed to assess the impact of SARS-CoV-2 infection on the number of NDTs and related medical and developmental outcomes in children. To compare continuous variables between those with and without SARS-CoV-2 infection, the Kruskal–Wallis test 15 was performed which helped determined whether the means of the two groups differed significantly. For categorical variables, chi-square test 16 was performed to evaluate differences in the frequency distributions between the two groups; this test assessed the likelihood of the observed differences to have occurred by chance. The distribution of normalized NDT values between groups was compared using the Kolmogorov–Smirnov (K–S) test. 17 This nonparametric method evaluates whether two samples differ significantly in their distributions. Additionally, chi-square tests were performed to examine the association between SARS-CoV-2 infection and prevalence of neurological diseases or disorders. To address the multiple comparisons across various diagnostic categories, the Benjamini–Hochberg procedure 18 was used, ensuring that the proportion of false positives among significant results remains below a specified threshold of 0.05.
In addition to the univariate analyses, we performed multivariate analysis to assess the effect of SARS-CoV-2 infection on the normalized NDTs, adjusting for potential confounders. As the outcome was a rate (i.e. the number of NDTs normalized by follow-up duration), treated as a continuous variable, we used linear regression to model the association. This approach eliminates the need for an offset term, which would be required in Poisson or negative binomial models, and enables direct interpretation of the absolute difference in visit rates. The dependent variable was the normalized NDT, and the independent variables included sex, any congenital disease (represented as a binary variable), normalized NDTs before the index date, SES (converted into dummy variables), and SARS-CoV-2 infection.
Results
The initial number of children in the database, before matching, was 242,745 (Table 1, Table S1). After excluding premature births and participants with missing information, the final study population was 227,728. Each child was followed up from birth until 1 January 2024, ensuring a comprehensive understanding of their development. The analysis revealed significant differences in the frequency of child development NDTs between children with and without SARS-CoV-2 infection. Among the 75,535 children with SARS-CoV-2 infection (33.2% of the study population), 416,642 NDTs were performed, averaging 5.52 NDTs per child. Conversely, 152,193 children without an infection (66.8% of the study population) underwent 501,457 NDTs, averaging 3.295 NDTs per child. A further breakdown indicated that among children with SARS-CoV-2 infection, 18,308 (24.2%) underwent at least one NDT. Among those without an infection, 26,741 (17.6%) underwent developmental NDTs. All subsequent analyses were performed on matched cohorts.
Demographic and clinical characteristics of the initial study and control groups before matching (values represent the entire study population over the study period).
Continuous variables are presented as mean (SD) and binary variables as numbers (%). p-values were calculated using the Kruskal–Wallis test for continuous variables and chi-square test for categorical variables.
SES: socioeconomic status; NDT: neurodevelopmental treatment.
Primary outcome of interest. Number of neurodevelopmental treatments following the index date served as the primary outcome. First, we considered all children, regardless of congenital disease status, to examine the broader implications of SARS-CoV-2 infection on the utilization of NDTs in children. Subsequently, for the sensitivity analysis, we excluded children with congenital diseases to isolate the impact of SARS-CoV-2 infection, thereby eliminating the confounding effects of pre-existing conditions.
We analyzed the impact of SARS-CoV-2 infection on the utilization of NDTs in children by comparing children with and without SARS-CoV-2 infection. Matches were identified for 64,311 children. The analysis revealed that among children with an SARS-CoV-2 infection, 11,519 (17.9%) underwent NDTs. In comparison, among those without an infection, 11,124 (17.3%) underwent NDTs (chi-square p < 1 × 10−5). We also compared the monthly average NDTs following the index date. Normalized NDT values were calculated by dividing the number of NDTs by the time in months for children with and without an SARS-CoV-2 infection. The dataset was divided into four groups: (a) children without an SARS-CoV-2 infection before the index date; (b) children with an SARS-CoV-2 infection before the index date; (c) children without an SARS-CoV-2 infection after the index date; and (d) children with an SARS-CoV-2 infection after the index date. A cumulative distribution function (CDF) represents the probability that a random variable takes on a value less than or equal to a specific value. In the presented graph (Figure 1), the CDFs for each group display the cumulative probabilities of the normalized NDT values, illustrating the likelihood of obtaining a value below or equal to a given point. Before the index date, the two groups of children showed no statistically significant difference in NDTs (K–S test, p = 1.0). In contrast, a noticeable difference was observed after the index date; children with an SARS-CoV-2 infection exhibited a statistically significantly higher number of normalized NDTs (K–S test p = 0.015). Excluding children with congenital diseases did not alter the outcome, with more treatments reported for the post-infection group (p = 3.43 × 10−3).

Data selection and filtration flow chart.
Multivariate analysis was conducted to account for potential confounding variables. The findings indicate a statistically significant positive association between SARS-CoV-2 infection and normalized NDTs following the index date (coefficient =0.0045 ± 0.002, p < 1 × 10−5), thereby reinforcing the robustness of the observed effect (Table 2).
Results of multivariate analysis using a linear regression model with normalized NDTs post-index date as the dependent variable.
NDT: neurodevelopmental treatments; SES: socioeconomic status; SARS-CoV-2: severe acute respiratory syndrome coronavirus 2; Coef: coefficient; CI: confidence interval.
NDTs comparisons stratified by subgroups. To further investigate the impact of SARS-CoV-2 infection, we conducted subgroup analyses to identify statistically significant differences between children with and without infection. All the results presented in this analysis reflect outcomes after the index date. The first subgroup analysis examined the age at which the child contracted SARS-CoV-2 infection. Significant increases in NDTs were observed after the index date among children aged 3–4 years (p = 4.0 × 10−4) and 4–5 years (p = 4.0 ×10−5).
The second subgroup analysis focused on SES. A significant difference was found between children with and without infection an SARS-CoV-2 infection in the low SES group, with p =6.23 × 10−5.
Finally, we analyzed the data based on the type of treatment needed. Among the different types of NDTs, general child development treatments (p = 0.038) and occupational therapy (p = 6.1 × 10−3) demonstrated statistically significant differences between children with and without infection after the index date.
2. Secondary outcomes of interest: Neurological and developmental disorders and other medical diagnoses comprised the secondary outcomes. We explored the association between SARS-CoV-2 infection and diagnosis of neurological and developmental diseases or disorders in children. All children with suitable matched controls were included, regardless of congenital disease status. The primary objective was to compare the prevalence of specific disorders between children who tested positive for SARS-CoV-2 infection and those who did not. Using chi-square tests, we compared the number of diagnoses between the groups with and without infection, focusing on disorders diagnosed before and after the index date. Additionally, we assessed the prevalence of these conditions before and after infection, using the date of SARS-CoV-2 diagnosis as the reference point for all children. Statistical significance was determined using chi-square test, and the Benjamini–Hochberg procedure
18
was performed to minimize the false discovery rate and account for multiple comparisons.
Pre-infection diagnoses. Our matching procedure successfully balanced the baseline characteristics between children with and without infection; however, several pre-existing conditions showed statistically significant differences between the groups before the index date. Most notably, ear, nose, and throat (ENT) diagnoses were substantially more prevalent in those with an SARS-CoV-2 infection (p = 1.9 × 10−5). We also observed modest but statistically significant differences in the odds ratios for eye problems, gastrointestinal conditions, injury-related conditions, and respiratory diseases (Appendix Figure S1).
Post-infection diagnoses. An analysis of diagnoses (Appendix Figure S2) revealed that several disorders were significantly associated with SARS-CoV-2 infection. ENT disorders exhibited statistically significant positive odds ratios and were more prevalent in children with an SARS-CoV-2 infection. Emotional difficulties, including anxiety disorders and acute stress reactions, were also statistically significantly more common in the group with an infection. Similarly, gastrointestinal conditions, such as esophageal reflux as well as headaches and respiratory diseases, including asthma and bronchitis, were more prevalent in those with an infection. In contrast, no significant differences were observed in neurodevelopmental diagnoses, except for autism spectrum disorder (ASD), which was more prevalent among children without an infection than among those with an SARS-CoV-2 infection.
Discussion
This study aimed to investigate the association between documented SARS-CoV-2 infection on neurodevelopmental outcomes in young children. We analyzed comprehensive population data collected both before and after exposure to SARS-CoV-2 infection, with the date of the first documented SARS-CoV-2 infection serving as the index date. Children without an SARS-CoV-2 infection were matched to this index date for comparison. Our findings indicate that children with SARS-CoV-2 infection are more likely to receive more NDTs, particularly occupational and related therapies, including speech and language therapy. However, no significant increase in neurodevelopmental diagnoses was observed. These results suggest that children with documented infection receive additional therapeutic interventions for the management of pre-existing developmental concerns. Documented SARS-CoV-2 infection reflects not only biological exposure but also the likelihood of testing and medical contact. Children with an SARS-CoV-2 infection demonstrated higher rates of non-neurodevelopmental medical diagnoses prior to infection, suggesting greater baseline engagement with healthcare services. Consequently, the observed increase in neurodevelopmental treatment utilization after the index date may reflect intensification or continuation of care among families already connected to developmental and medical services, rather than a direct effect of SARS-CoV-2 infection itself. This interpretation is supported by the absence of a post-infection increase in new neurodevelopmental diagnoses.
Several studies have reported increased rates of language and communication delays among children born or raised during the COVID-19 pandemic.2–9 In many cases, it is unclear if exposure to and illness due to SARS-CoV-2 infection or environmental changes during the pandemic, to some extent, influenced children’s health and development. The pandemic has led to pivotal lifestyle changes with possible mental and social alterations in children and parents. To examine the effect of social disadvantages on health and development during the pandemic, Weyes and Rigó conducted a study involving 800 children; 19 health and development deteriorated across all social statuses; however, the effect was more pronounced in more deprived children. During the pandemic, changes in sleep patterns, decreased physical activity, increased screen time, and lockdowns could have potentially impaired children’s global development.20–23 Parents, especially with young children, experienced significant mental health symptoms during the outbreaks. Higher rates of parent depression, anxiety, and stress as well as higher alcohol consumption were reported. 24 Employment status, family conflicts, and financial burden all contribute to increased parental mental distress, which might be a risk factor for a child’s development.25,26 Importantly, most of these studies examined the effects of pandemic-related environmental exposures rather than documented SARS-CoV-2 infection in the child. These two exposures are conceptually distinct. The present study specifically evaluates the association between documented SARS-CoV-2 infection and subsequent utilization of neurodevelopmental treatments and diagnoses. Therefore, the absence of an increase in new neurodevelopmental diagnoses following infection does not contradict pandemic-era findings as many children included in those studies had never contracted SARS-CoV-2 infection.
Studies that compared children with an SARS-CoV-2 infection and healthy controls10,11 included a small number of participants and higher proportion of participants with developmental delays in the group of children with an SARS-CoV-2 infection. We examined the population-level data and followed the same cohort of children before and after documented SARS-CoV-2 infection; in our study, the incidence of developmental delay was not more in children with an infection. The divergence observed between increased neurodevelopmental treatment utilization and stable rates of formal neurodevelopmental diagnoses is a central finding of this study. Neurodevelopmental treatments may be initiated in response to subthreshold developmental concerns, parental anxiety, or transient delays that do not ultimately result in a formal diagnosis. This may be particularly relevant in the context of the COVID-19 pandemic, during which heightened societal concern regarding child development may have lowered the intervention thresholds. In the absence of standardized developmental assessment scores, treatment utilization should be interpreted as a measure of perceived developmental need and service engagement rather than a direct proxy for neurodevelopmental impairment. The absence of an increase in new neurodevelopmental diagnoses following infection suggests that the increased treatment utilization reflects enhanced monitoring and support rather than underlying pathology.
Although we were unable to determine the exact cause for the lower number of new ASD diagnoses observed after the index date, we did not observe any increase in ASD diagnoses among children with documented SARS-CoV-2 infection.
Children with an SARS-CoV-2 infection exhibited a higher incidence of ear infections, respiratory illnesses, and gastrointestinal disorders before the SARS-CoV-2 infection, which may have increased their underlying vulnerability to subsequent infections. It is also possible that the presence of underlying medical issues led parents to seek medical care more frequently for these children. However, children with and without infection were matched for pre-index neurodevelopmental treatment utilization, and no differences were observed prior to infection. Matching on congenital diseases and pre-index neurodevelopmental treatment utilization, along with sensitivity analyses excluding congenital conditions, suggests that the observed post-infection increase in neurodevelopmental treatment utilization is not explained by baseline medical comorbidity.
Special attention should be given to the long-term emotional well-being of these children, ensuring that they receive adequate care and support.
The findings were robust. In sensitivity analyses, we excluded children with congenital diseases or changed the hyperparameters for matching, and the results remained stable.
Certain study limitations should be considered. Our source for SARS-CoV-2 infection data was physician’s diagnosis recorded in the patient’s electronic records, wherein data regarding the detection method (e.g. polymerase chain reaction (PCR) or antigen test) was missing. Therefore, mild or asymptomatic infections may have been missed, leading to nondifferential misclassification that would bias results toward the null. Differential misclassification may also have occurred if children with greater healthcare utilization were more likely to have infections documented. To address this, children with and without an infection were matched on pre-index neurodevelopmental treatment utilization, and no differences were observed prior to infection. Nevertheless, the possibility of residual misclassification cannot be ruled out and should be considered when interpreting the results.
Furthermore, we did not have access to data on any additional treatments a child may have received in private healthcare settings. However, we believe that this limitation did not substantially impact the results. We also did not have information regarding infection severity. Although previous studies have suggested that the long-term effects of SARS-CoV-2 infection are associated with more severe infection, 27 this is unlikely to affect our results as majority of SARS-CoV-2 infections in young children are mild in intensity. 28 Furthermore, we believe that the approximately 2-year follow-up duration was insufficient to comprehensively estimate the long-term effects of SARS-CoV-2 infection.
Standardized neurodevelopmental assessment scores were not available in the EMRs because developmental evaluations are conducted using heterogeneous instruments and are not recorded in a structured, extractable format. Therefore, neurodevelopmental treatment utilization was used as a clinically relevant proxy for developmental concern and service need.
We acknowledge that the analysis did not explicitly adjust for external events such as lockdowns, service closures, and other pandemic-related disruptions. However, because controls were matched to cases by date of birth and assigned identical index dates, both groups were similarly exposed to these systemic events over time. Furthermore, as Israel is a relatively small country where public health measures, including lockdowns and service modifications, were implemented at the national level rather than regionally, the impact of these events is expected to have been uniform across the study population. Therefore, any influence of such disruptions would likely be nondifferential between groups and not introduce systematic bias in the comparison of NDT rates.
Conclusions
Children with documented SARS-CoV-2 infection had a significantly higher rate of NDTs post-infection compared with those without an infection. However, we did not observe an increase in the number of neurodevelopmental diagnoses among children with an SARS-CoV-2 infection. Although matching included prior neurodevelopmental treatment utilization, which reflects baseline developmental concerns and healthcare engagement, residual confounding from unmeasured parental, psychosocial, and healthcare access factors is likely and should be considered when interpreting the results. Future studies on such cohorts should investigate additional diagnoses, such as specific learning disabilities and attention deficit hyperactivity disorder, which are typically identified later in childhood and adolescence. Additionally, family-based or sibling-comparison designs may help distinguish the biological effects of infection from shared familial and healthcare-seeking factors.
Supplemental Material
sj-pdf-1-imr-10.1177_03000605261436920 - Supplemental material for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study
Supplemental material, sj-pdf-1-imr-10.1177_03000605261436920 for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study by Davidovitch Michael, Leventer Matan, Patalon Tal and Rappoport Nadav in Journal of International Medical Research
Supplemental Material
sj-pdf-2-imr-10.1177_03000605261436920 - Supplemental material for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study
Supplemental material, sj-pdf-2-imr-10.1177_03000605261436920 for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study by Davidovitch Michael, Leventer Matan, Patalon Tal and Rappoport Nadav in Journal of International Medical Research
Supplemental Material
sj-pdf-3-imr-10.1177_03000605261436920 - Supplemental material for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study
Supplemental material, sj-pdf-3-imr-10.1177_03000605261436920 for Neurodevelopmental treatment needs after severe acute respiratory syndrome coronavirus 2 infection in young children: A cohort study by Davidovitch Michael, Leventer Matan, Patalon Tal and Rappoport Nadav in Journal of International Medical Research
Footnotes
Acknowledgments
None.
Author contributions
MD conceived and initiated the study, contributed to study design, conducted data analysis, and led manuscript writing and revision.
ML was responsible for data retrieval, statistical analysis, and interpretation of the results.
TP contributed to the study conception and critically reviewed and edited the manuscript.
NR contributed to data analysis, statistical interpretation, and manuscript writing and revision.
Data availability statement
The data used in this study contain sensitive personal information and cannot be shared publicly to protect participants' privacy. Deidentified data may be made available by the corresponding author on reasonable request and subject to appropriate ethical approvals.
Declaration of conflicting interests
The authors declare no conflicts of interest.
Ethical approval
The study received ethical approval from the Ethical Committee of Maccabi Healthcare Services (Approval no. 0060-22-MHS).
Funding
A grant from the Maccabi Research and Innovation Center supported this study; ML was partially funded by grant of the Maccabi Research and Innovation Center. TP was part of the grant committee at the Maccabi Research and Innovation Center.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
