This cohort study evaluates the association of premature birth and other child and maternal factors with school readiness in children born in Manitoba, Canada.
Key Points
Question
Does prematurity affect school readiness in a population-based cohort of children?
Findings
In this cohort study of 63 277 children in kindergarten at the time of data collection, children born preterm were more likely to be not school ready when assessed in kindergarten compared with children born full term (35% vs 28%), a statistically significant difference. Similar differences between these 2 groups were found in all 5 domains of school readiness, but a difference in school readiness was not observed between children born preterm and their siblings born full term.
Meaning
Findings of this study suggest that prematurity and other child and maternal factors were associated with lack of school readiness in this population-based cohort of children.
Abstract
Importance
Children born preterm may experience learning challenges at school. However, there is a paucity of data on the school readiness of these children as they prepare to begin grade 1.
Objective
To examine the association between prematurity and school readiness in a population-based cohort of children.
Design, Setting, and Participants
This cohort study was conducted in the province of Manitoba, Canada, and involved 2 cohorts of children in kindergarten at the time of data collection. The population-based cohort included children born between January 1, 2000, and December 31, 2011, whose school readiness was assessed in kindergarten using the Early Development Instrument (EDI) data. The sibling cohort comprised children born preterm and their closest-in-age siblings born full term. Data were analyzed between March 12 and September 28, 2021.
Exposures
Preterm birth, defined as gestational age (GA) less than 37 weeks.
Main Outcomes and Measures
The primary outcome was vulnerability in the EDI, defined as a score below the tenth percentile of the Canadian population norms for any 1 or more of the 5 EDI domains (physical health and well-being, social competence, emotional maturity, language and cognitive development, and communication skills and general knowledge). Logistic regression models were used to identify the factors associated with vulnerability in the EDI. P values were adjusted for multiplicity using the Simes false discovery method.
Results
Of 86 829 eligible children, 63 277 were included, of whom 4352 were preterm (mean [SD] GA, 34 [2] weeks; 2315 boys [53%]) and 58 925 were full term (mean [SD] GA, 39 (1) weeks; 29 885 boys [51%]). Overall, 35% of children (1536 of 4352) born preterm were vulnerable in the EDI compared with 28% of children (16 449 of 58 925) born full term (adjusted odds ratio [AOR], 1.32; 95% CI, 1.23-1.41; P < .001]). Compared with children born full term, those born preterm had a higher percentage of vulnerability in each of the 5 EDI domains. In the population-based cohort, prematurity (34-36 weeks’ GA: AOR, 1.23 [95% CI, 1.14-1.33]; <34 weeks’ GA: AOR, 1.72 [95% CI, 1.48-1.99]), male sex (AOR, 2.24; 95% CI, 2.16-2.33), small for gestational age (AOR, 1.31; 95% CI, 1.23-1.40), and various maternal medical and sociodemographic factors were associated with EDI vulnerability. In the sibling cohort, EDI outcomes were similar for both children born preterm and their siblings born full term except for the communication skills and general knowledge domain (AOR, 1.39; 95% CI, 1.07-1.80) and Multiple Challenge Index (AOR, 1.43; 95% CI, 1.06-1.92), whereas male sex (AOR, 2.19; 95% CI, 1.62-2.96) and maternal age at delivery (AOR, 1.53; 95% CI, 1.38-1.70) were associated with EDI vulnerability.
Conclusions and Relevance
Results of this study suggest that, in a population-based cohort, children born preterm had a lower school-readiness rate than children born full term, but this difference was not observed in the sibling cohort. Child and maternal factors were associated with lack of school readiness among this population-based cohort.
Introduction
School readiness refers to a set of developmental and behavioral skills that children require to transition successfully to the school environment.1 These skills are foundational for children’s subsequent learning and are a factor in their future academic achievement, social well-being, and health as adults.2,3,4,5 Thus, school readiness is a major milestone during early childhood, providing an opportunity to identify and support children, especially those at risk for educational challenges.
Children born preterm are at risk for long-term neurodevelopmental challenges, including learning difficulties.6,7,8 However, the few studies that evaluated the early educational performance, including school readiness, of these children were limited by small cohort sizes, sampling biases, use of nonvalidated clinic-based evaluation tools, residual sociodemographic confounding, and scant assessments of school readiness.9,10,11,12,13,14 In this study, we aimed to examine the association between prematurity and school readiness in a population-based cohort of children. We hypothesized that children born preterm have a lower rate of school readiness compared with children born full term.
Methods
This retrospective cohort study used the Population Research Data Repository at the Manitoba Centre for Health Policy, University of Manitoba, Canada. The repository contains linked, populationwide data.15 Approvals for this study were obtained from the University of Manitoba Health Research Ethics Board, Manitoba Health Information Privacy Committee, and other data custodians. The University of Manitoba Health Research Ethics Board also determined that consent from participants was not required because this was a retrospective study.
We used hospital abstracts, social assistance data, physician visit records, prescription medication records, a population health registry, educational outcomes data, the Diabetes Education Resource for Children and Adolescents database, Early Development Instrument (EDI) data, Child and Family Services data, and data from a newborn-mother infant risk screening program (Families First Screen). Race and ethnicity data were not collected. Socioeconomic status was assessed using an area-level indicator derived from the Canadian census called the Socioeconomic Factor Index-Version 2 (SEFI-2), for which higher values indicate lower socioeconomic status (eTable 1 in the Supplement).16 A scrambled personal health identifier was used to link across data sets. These data have virtually complete population coverage given that the province of Manitoba has publicly funded universal health care.17 Validated mother-child linkages in the repository were used in the study cohorts.18,19
Using the repository data, we created 2 cohorts of children who were in kindergarten at the time of data collection: a population-based cohort and a sibling cohort. The primary analysis, determined a priori, focused on a population-based cohort of children born in Manitoba between January 1, 2000, and December 31, 2011, who had continuous provincial health insurance coverage from birth until school entry and underwent school-readiness assessment in kindergarten. We excluded children who (1) had a substantial congenital anomaly (based on the Congenital Anomalies in Canada 2013 report20), (2) could not be linked to maternal records or school-readiness outcomes, and (3) had mothers without provincial health insurance coverage at least 2 years before the children’s birth. Gestational age (GA) in completed weeks was obtained from hospital delivery records and was based on maternal last menstrual period or second-trimester ultrasonographic dating when last menstrual period was unavailable or inaccurate, per local guidelines. Preterm birth (<37 weeks GA) was the exposure of interest, and full-term birth (≥37 weeks) was considered to be unexposed.
The secondary analysis focused on a sibling cohort, a subset of the population-based cohort. The sibling cohort comprised children born preterm and their siblings closest in age who were born full term.
Manitoba teachers complete the EDI questionnaire every other year for their students across all 37 public school divisions during the latter half of the kindergarten year (February-March). The EDI data available for the following years were included in the study: 2005 to 2006, 2006 to 2007, 2008 to 2009, 2010 to 2011, 2012 to 2013, 2014 to 2015, and 2016 to 2017.
Outcomes
Outcomes were derived from the EDI, a comprehensive, 103-item questionnaire21 that is widely used as a population-based, validated measure of school readiness and has excellent psychometric properties.22,23,24 It assesses 5 developmental domains: (1) physical health and well-being, (2) emotional maturity, (3) social competence, (4) language and cognitive development, and (5) communication skills and general knowledge. Performance in each domain is expressed as a continuous score and as a percentile based on Canadian population norms. Because EDI domains measure different aspects of children’s development, the domain scores are not summed. Instead, children are considered to be vulnerable in the EDI if they score below the tenth percentile of the Canadian population norms for any 1 or more of the EDI domains.
The primary outcome was vulnerability in the EDI. Secondary outcomes were vulnerability in each EDI domain; vulnerability in 1, 2, 3, 4, or 5 domains; scores in each EDI domain; the Multiple Challenge Index, defined as vulnerability in 3 or more EDI domains; and need for additional support in kindergarten.
Statistical Analysis
Descriptive statistics were used to summarize baseline covariates and outcomes. Standardized mean differences were used to compare baseline covariates, and differences greater than 0.1 were considered to be substantial. To compare outcomes, we used unpaired, 2-tailed t test for continuous variables and χ2 test for categorical variables. The primary analysis used multivariable logistic regression in the population-based cohort to assess whether prematurity was associated with vulnerability in the EDI after adjusting for potential confounders. Maternal and neonatal covariates available in the perinatal period were chosen from existing literature and were included in the model.10,12,13,14 A list of these variables and their definitions are included in eTable 1 in the Supplement. We used E-value to measure unmeasured confounding in the models.25 For the secondary outcomes, we reported the unadjusted analyses only.
Within the preterm group, we performed a subgroup analysis by classifying individuals a priori into 3 GA groups: less than 28 weeks, 28 to 33 weeks, and 34 to 36 weeks. Post hoc, because of small numbers in the lower GA categories, we had to combine them in regression models for both population-based and sibling cohorts.
For the sibling analysis, a conditional logistic regression model was fit to the data to account for clustering of siblings within family units. We evaluated for multicollinearity among independent model variables using the variance inflation factor; none met the criterion of variance inflation factor greater than 10. To adjust for P values of independent variables resulting from multiple comparisons in the regression models, we used the Simes step-up method to control the false discovery rate.26 A 2-sided P < .05 was considered to be significant.
Statistical analyses were performed with SAS, version 9.4 (SAS Institute Inc). Data were analyzed between March 12 and September 28, 2021.
Results
Population-Based Cohort
Of the 86 829 eligible children, 63 277 children (4352 children born preterm; 58 925 children born full term) were included and formed the population-based cohort (Figure 1). Baseline characteristics of children who were excluded because they were not linkable to EDI data (n = 21 114) and children who were included in the population-based cohort are provided in eTable 2 in the Supplement. Maternal age at delivery and age at first childbirth were higher and birth order and SEFI-2 scores were lower among the included mothers compared with the excluded mothers.
Figure 1. Flowchart of Patients.
EDI indicates Early Development Instrument.
There were differences in baseline characteristics between children born preterm and full term in the population-based cohort (Table 1). The 4352 children born preterm had a mean (SD) GA of 34 (2) weeks and included 2315 boys (53%) and 2037 girls (47%). The 58 925 children born full term had a mean (SD) GA of 39 (1) weeks and included 29 885 boys (51%) and 29 040 girls (49%). Children born preterm were less likely than children born full term to be born to mothers in a rural residence and were more likely to be born by cesarean delivery and to mothers with diabetes, with prenatal psychological distress, who were smokers, or who were receiving income assistance at the time of delivery. A higher percentage of children born preterm entered Child and Family Services care between birth and EDI assessment than children born full term. In the population-based cohort compared with the sibling cohort, mean (SD) maternal age at delivery and birth order were higher, but the mean (SD) 5-minute Apgar score was lower among children born preterm vs children born full term.
Table 1. Baseline Characteristics of the Population-Based and Sibling Cohorts.
| Characteristic | Population-based cohort | Sibling cohort | ||||
|---|---|---|---|---|---|---|
| Children born preterm, No. (%) (n = 4352) | Children born full term, No. (%) (n = 58 925) | Standardized mean difference (95% CI) | Children born preterm, No. (%) (n = 1296) | Siblings born full term, No. (%) (n = 1296) | Standardized mean difference (95% CI) | |
| Maternal | ||||||
| Age at delivery, y | ||||||
| Median (IQR) | 29 (24 to 33) | 28 (24 to 32) | NA | 28 (24 to 32) | 28 (23 to 32) | NA |
| Mean (SD) | 29 (6) | 28 (6) | −0.77 (−0.94 to −0.59) | 28 (5) | 28 (5) | −0.44 (−0.86 to −0.02) |
| Age at first childbirth, y | ||||||
| Median (IQR) | 24 (19 to 29) | 24 (20 to 29) | NA | 23 (19 to 28) | 23 (19 to 28) | NA |
| Mean (SD) | 25 (6) | 25 (6) | −0.13 (−0.31 to 0.04) | 23 (5) | 23 (5) | 0.00 (−0.42 to 0.42) |
| Rural residence | 1641 (38) | 24 047 (41) | NA | 558 (43) | 551 (43) | NA |
| Birth order | ||||||
| Median (IQR) | 2 (1 to 3) | 2 (1 to 3) | NA | 2 (1 to 3) | 2 (1.5 to 3) | NA |
| Mean (SD) | 2 (1) | 2 (1) | −0.13 (−0.16 to −0.09) | 3 (2) | 2 (1) | −0.14 (−0.25 to −0.02) |
| Diabetes | 325 (7) | 1209 (2) | NA | 66 (5) | 52 (4) | NA |
| Prenatal psychological distress | 488 (11) | 5495 (9) | NA | 136 (10) | 133 (10) | NA |
| Smoking | 843 (19) | 9565 (16) | NA | 297 (23) | 276 (21) | NA |
| Substance use disorder | 164 (4) | 1589 (3) | NA | 54 (4) | 45 (3) | NA |
| Receipt of income assistance | 698 (16) | 7461 (13) | NA | 266 (21) | 283 (22) | NA |
| SEFI-2 score | ||||||
| Median (IQR) | 0.1 (−0.5 to 0.7) | 0.0 (−0.5 to 0.6) | NA | 0.1 (−0.4 to 0.8) | 0.1 (−0.4 to 0.9) | NA |
| Mean (SD) | 0.2 (1.0) | 0.1 (1.0) | −0.09 (−0.12 to −0.06) | 0.3 (1.0) | 0.3 (1.1) | 0.02 (−0.06 to 0.10) |
| Married or in a common-law relationship | 1520 (35) | 23 258 (39) | NA | 441 (34) | 458 (35) | NA |
| Cesarean delivery | 1564 (36) | 11 458 (19) | NA | 435 (34) | 260 (20) | NA |
| Neonatal | ||||||
| Gestational age, wk | ||||||
| Median (IQR) | 35 (34 to 36) | 40 (39 to 40) | NA | 35 (34 to 36) | 39 (38 to 40) | NA |
| Mean (SD) | 34 (2) | 39 (1) | 4.98 (4.94 to 5.02) | 35 (2) | 39 (1) | 4.25 (4.12 to 4.37) |
| Birth weight, g | ||||||
| Median (IQR) | 2540 (2122 to 2910) | 3528 (3212 to 3859) | NA | 2591 (2200 to 2942) | 3421 (3065 to 3765) | NA |
| Mean (SD) | 2512 (663) | 3543 (491) | 1030 (1015 to 1046) | 2572 (626) | 3428 (517) | 856 (811 to 900) |
| Female sex | 2037 (47) | 29 040 (49) | NA | 575 (44) | 650 (50) | NA |
| Male sex | 2315 (53) | 29 885 (51) | NA | 721 (56) | 646 (50) | NA |
| 5-min Apgar score | ||||||
| Median (IQR) | 9 (8 to 9) | 9 (9 to 9) | NA | 9 (8 to 9) | 9 (9 to 9) | NA |
| Mean (SD) | 8 (1) | 9 (1) | 0.40 (0.38 to 0.42) | 9 (1) | 9 (1) | 0.33 (0.28 to 0.39) |
| SGA | 369 (8) | 4410 (7) | NA | 96 (7) | 120 (9) | NA |
| Taken to care by Child and Family Services | 407 (9) | 2850 (5) | NA | 165 (13) | 154 (12) | NA |
| Siblings born full term older than children born preterm | NA | NA | NA | 598 (46) | 598 (46) | 598 (46) |
| Age difference between children born preterm and sibling born full term, median (IQR), y | NA | NA | NA | 1 (−2 to 3) | 1 (−2 to 3) | 1 (−2 to 3) |
Abbreviations: NA, not applicable; SEFI-2, Socioeconomic Factor Index-Version 2; SGA, small for gestational age.
The adjusted outcomes in the population-based cohort showed that 35% of children (1536 of 4352) born preterm were vulnerable compared with 28% of children (16 449 of 58 925) born full term (adjusted odds ratio [AOR], 1.32; 95% CI, 1.23-1.41; P < .001]) (Table 2). These results showed a dose-response association among children born preterm, with a higher percentage of those born at earlier GAs (<28 weeks or 28-33 weeks) being vulnerable in any EDI domain (48% [43 of 89] and 40% [315 of 796]), each of the 5 EDI domains (eg, physical health and well-being: 34% [30 of 89] vs 22% [172 of 796]), Multiple Challenge Index (25% [22 of 89] vs 17% [136 of 796]), and need for additional support in kindergarten (22% [20 of 89] vs 5% [36 of 796]) compared with those born at later GAs (34-36 weeks). A similar pattern was observed for the scores in each individual domain of the EDI (eFigure and eTable 3 in the Supplement). Physical health and well-being, language and cognitive development, and communication skills and general knowledge were the most affected domains among children born preterm, whereas social competence and emotional maturity were the least affected domains.
Table 2. Results for the Population-Based Cohort.
| Characteristic | Gestational age, No. of participants (%) | AOR (95% CI) | P valuea,b | ||||
|---|---|---|---|---|---|---|---|
| <28 wk (n = 89) | 28-33 wk (n = 796) | 34-36 wk (n = 3467) | Preterm (n = 4352) | Full term (n = 58 925) | |||
| Age at EDI assessment, mean (SD), y | 4.8 (0.5) | 4.7 (0.5) | 4.7 (0.5) | 4.7 (0.5) | 4.7 (0.5) | NA | NA |
| Outcomes | |||||||
| Vulnerability in any EDI domain | 43 (48) | 315 (40) | 1178 (34) | 1536 (35) | 16 449 (28) | 1.32 (1.23-1.41) | <.001 |
| Vulnerability in the EDI domain of | |||||||
| Physical health and well-being | 30 (34) | 172 (22) | 608 (18) | 810 (19) | 7623 (13) | 1.40 (1.29-1.53) | <.001 |
| Language and cognitive development | 27 (30) | 167 (21) | 572 (17) | 766 (18) | 6794 (12) | 1.50 (1.37-1.64) | <.001 |
| Social competence | 23 (26) | 148 (19) | 496 (14) | 667 (15) | 6856 (12) | 1.25 (1.14-1.37) | <.001 |
| Emotional maturity | 15 (17) | 139 (18) | 505 (15) | 659 (15) | 6817 (12) | 1.27 (1.16-1.40) | <.001 |
| Communication skills and general knowledge | 25 (28) | 156 (20) | 520 (15) | 701 (16) | 6223 (11) | 1.50 (1.37-1.64) | <.001 |
| Multiple Challenge Index | 22 (25) | 136 (17) | 443 (13) | 601 (14) | 5032 (9) | 1.54 (1.40-1.70) | <.001 |
| Need for additional support in kindergarten | 20 (22) | 36 (5) | 94 (2) | 150 (3) | 980 (2) | 1.88 (1.57-2.24) | <.001 |
Abbreviations: AOR, adjusted odds ratio; EDI, Early Development Instrument; NA, not applicable.
Between children born preterm and children born full term.
P values were adjusted using the Simes false discovery rate method and were adjusted for neonate sex and small-for-gestational-age status; maternal age at delivery, age at first childbirth, place of residence, birth order, diabetes, prenatal psychological distress, smoking, substance use disorder, receipt of income assistance at time of delivery, and SEFI-2 (Socioeconomic Factor Index–Version 2) score; and year of birth.
A higher percentage of children born preterm were vulnerable in 1, 2, 3, 4, or 5 EDI domains than children born full term in the population-based cohort. The GA gradient was seen for vulnerability in 2 or more domains of the EDI as shown in Figure 2. Specifically, 12% of children born at less than 28 weeks’ gestation, 9% born between 28 and 33 weeks’ gestation, and 7% born between 34 and 36 weeks’ gestation were vulnerable in 2 domains, whereas 8% of children born less than 28 weeks’ gestation, 7% born between 28 and 33 weeks’ gestation, and 5% born between 34 and 36 weeks’ gestation were vulnerable in 3 domains. In addition, 37% of children born at less than 28 weeks’ gestation, 26% born between 28 and 33 weeks’ gestation, and 20% born between 34 and 36 weeks’ gestation were vulnerable in more than 1 domain compared with 16% of children born at 37 weeks’ gestation or longer.
Figure 2. Vulnerability in Early Development Instrument (EDI) Domains for the Population-Based Cohort and Sibling Cohort.
GA indicates gestational age.
The regression model of the population-based cohort is shown in eTable 5 in the Supplement. Prematurity (34-36 weeks: AOR, 1.23 [95% CI, 1.14-1.33]; <34 weeks: AOR, 1.72 [95% CI, 1.48-1.99]); male sex (AOR, 2.24; 95% CI, 2.16-2.33); small for gestational age (SGA; AOR, 1.31 [95% CI, 1.23-1.40]); higher birth order (AOR, 1.13; 95% CI, 1.10-1.16); and maternal diabetes (AOR, 1.13; 95% CI, 1.10-1.16), smoking (AOR, 1.43; 95% CI, 1.36-1.50), prenatal psychological distress (AOR, 1.17; 95% CI, 1.10-1.24), receipt of income assistance (AOR, 1.76; 95% CI, 1.66-1.86), and higher SEFI-2 score (AOR, 1.26; 95% CI, 1.24-1.29) were the factors associated with being vulnerable in the EDI. In contrast, higher maternal age at delivery and first childbirth as well as rural residence were the factors associated with not being vulnerable in the population-based cohort. The E-value for unmeasured confounding for GA of less than 34 weeks was 1.95 in the population-based model, with a lower 95% CI of 1.73.
Sibling Cohort
Of the 4352 children born preterm, only 1296 had a sibling born full term (median [IQR] GA, 35 [34-36] weeks; 721 boys [56%], 575 girls [44%]) and were included in the sibling cohort. The cohort also included 1296 siblings born full term (median [IQR] GA, 39 [38-40] weeks; 646 boys [50%], 650 girls [50%]). In the sibling cohort, maternal median (IQR) age at delivery (28 [24-32] years vs 28 [23-32] years), male sex (721 [56%] vs 646 [50%]), and birth by cesarean delivery (435 [34%] vs 260 [20%]) were higher but median (IQR) 5-minute Apgar score (9 [8-9] vs 9 [9-9]) was lower among children born preterm compared with their siblings born full term.
The adjusted primary and secondary outcomes were similar between children born preterm and their siblings born full term except for vulnerability in the communication skills and general knowledge EDI domain (AOR, 1.39; 95% CI, 1.07-1.80) and Multiple Challenge Index (AOR, 1.43; 95% CI, 1.06-1.92) (Table 3). The differences in their scores were seen in the EDI domains of communication skills and general knowledge as well as language and cognitive domains (eTable 4 in the Supplement). Only the presence of vulnerability in all 5 domains was different between children born preterm and their siblings born full term (Figure 2). For example, 5% of children born preterm (65 of 1296) were vulnerable in all 5 domains compared with 2% of their siblings born full term (26 of 1296). Male sex (AOR, 2.19; 95% CI, 1.62-2.96) and maternal age at delivery (AOR, 1.53; 95% CI, 1.38-1.70) were the only factors associated with vulnerability in the EDI (eTable 5 in the Supplement).
Table 3. Results for the Sibling Cohort.
| Characteristic | Participants, No. (%) | AOR (95% CI) | P valuea | |
|---|---|---|---|---|
| Children born preterm (n = 1296) | Siblings born full term (n = 1296) | |||
| Age at EDI assessment, mean (SD), y | 4.7 (0.5) | 4.7 (0.5) | NA | NA |
| Outcomes | ||||
| Vulnerability in any EDI domain | 462 (36) | 451 (35) | 1.02 (0.83-1.25) | .94 |
| Vulnerability in the domain of | ||||
| Physical health and well-being | 255 (20) | 227 (18) | 1.19 (0.93-1.53) | .21 |
| Language and cognitive development | 232 (18) | 202 (15) | 1.32 (1.00-1.74) | .07 |
| Social competence | 196 (15) | 182 (14) | 1.00 (0.76-1.32) | .99 |
| Emotional maturity | 186 (15) | 171 (13) | 1.04 (0.79-1.37) | .91 |
| Communication skills and general knowledge | 240 (19) | 195 (15) | 1.39 (1.07-1.80) | .02 |
| Multiple Challenge Index | 190 (15) | 156 (12) | 1.43 (1.06-1.92) | .03 |
| Need for additional support in kindergarten | 39 (3) | 29 (2) | 1.36 (0.63-2.93) | .53 |
Abbreviations: AOR, adjusted odds ratio; EDI, Early Development Instrument; NA, not applicable.
P values were adjusted using the Simes false discovery rate method and were adjusted for neonate sex and small-for-gestational-age status; maternal age at delivery, age at first childbirth, place of residence, birth order, diabetes, prenatal psychological distress, smoking, substance use disorder, receipt of income assistance at time of delivery, and SEFI-2 (Socioeconomic Factor Index–Version 2) score; and year of birth.
Discussion
In this cohort study of school readiness among a population-based cohort of children, we found that, after adjusting for a variety of potentially confounding variables, children born preterm had lower rates of school readiness than children born full term. Furthermore, there was a clear GA gradient, with children born at earlier GAs (<28 weeks or 28-33 weeks) having lower rates of school readiness compared with those born at later GAs (34-36 weeks). The population-based regression model identified prematurity, male sex, SGA, and various maternal medical and sociodemographic characteristics as factors associated with lack of school readiness, whereas male sex and maternal age at delivery were the only factors with such an association in the sibling model.
Children born preterm are particularly vulnerable to educational challenges at school.8 Various factors, including prematurity itself, its complications, and the postnatal exposure of the preterm brain to noxious stimuli during the stay in the neonatal intensive care unit, can disrupt the rapid brain growth in these children. Functionally, these disruptions can play a role in impairments in cognitive abilities (including IQ), neurophysiological processes, executive functions, and behavioral problems (including impulsivity and attention deficit), all of which can affect school performance.27,28
Contrary to the notion that children born preterm with early school challenges would catch up to their peers over time, studies found that school challenges persist and may worsen in the middle and high school years as educational demands increase.29,30 The Matthew Effect, a phenomenon wherein children with deficits in 1 educational domain often perform worse over time unless they receive adequate support and subsequently progress to general intellectual and cognitive disabilities, could also have implications for children’s school performance.31 These findings underscore the importance of evaluating children born preterm during their early years.
The concept of school readiness emerged in the 1990s after the US National Education Goals Panel recommended that “every child in the US be school ready by the year 2000.”32 The American Academy of Pediatrics endorsed these recommendations and defined the school readiness framework to facilitate children’s school transition and support those who have challenges.1 School readiness is an important outcome associated with children’s future educational achievement and is amenable to remedial interventions.33,34,35,36 For example, the study by D’Anguilli et al36 found that vulnerability in the EDI (<tenth percentile in ≥1 domain) was indicative of poor performance in grade 4–level reading, writing, and math in their population-based cohort. D’Anguilli et al36 also noted a direct correlation between the number of domains in which children were vulnerable and the rates of poor performance in their study. Despite the assessment of multiple domains of a child’s development as part of school readiness, there is some evidence to suggest that reading, math, and attention skills are most associated with their future academic success.34
Although limited, previous studies have reported that children born preterm were more likely to be not school ready than children born full term. Comparing the results of the present study with previous research is challenging because of the variability in patient selection, age at assessment, and type of assessment tools used.9,10,11,12,13,14 Roberts et al12 evaluated school readiness among a cohort of very preterm neonates (mean [SD] GA, 27.4 [1.9] weeks) at 5 years corrected age using multiple assessment tools. Overall, 72% of the cohort in the study was considered to be not school ready, whereas 44% were not ready in more than 1 domain of school readiness. The proportion of children with vulnerability ranged from 20% to 53% across the domains in the Roberts et al12 cohort. In comparing the previous cohort with children born less than 28 weeks’ gestation in the present study, we observed that 48% of these children were considered to be vulnerable in any EDI domain and 37% in more than 1 EDI domain. Similarly, vulnerability ranged from 17% to 34% in the population-based cohort, with the percentages of vulnerability in physical health and well-being as well as language and cognitive EDI domains being similar to those reported in previous studies, whereas percentages of vulnerability in communication skills and social competence domains were different. Patrianakos-Hoobler et al10 found that 33% of neonates born preterm with respiratory distress syndrome (mean [SD] GA, 27.5 [2.6] weeks) were not school ready. Using only reading and math performance in kindergarten, Shah et al13 found that a higher proportion of children born very preterm (8.3%-12.8%) were not school ready compared with the more mature children born preterm (32-36 weeks GA: 5%-6%; full term: approximately 3.5%-4%), confirming the dose-response association of GA found in the present study.
The results of this study concur with the findings from existing studies that have identified risk factors, such as male sex, low socioeconomic status, lower maternal education, Black race, younger age at assessment, and lack of preschool experience, for not being school ready in this population.10,12,13,14 Also, we found additional factors, such as SGA, higher birth order, and maternal psychological distress, smoking, and diabetes, to be associated with lack of school readiness. In a meta-analysis, SGA was shown to be associated with poor neurodevelopmental outcomes that included motor, language, cognitive, and behavioral domains among children born full term and those born preterm.37 Altered growth of the frontal lobes and hippocampus during the intrauterine period, which are areas of the brain related to learning capabilities and executive functions, might be a factor in these findings.38,39,40 Similarly, previous studies have reported the adverse implications of maternal depression for school readiness and diabetes for later school performance in children.41,42,43 In contrast, rural residence, increasing maternal age at delivery, and first childbirth seemed to be protective factors in the population-based cohort.
The extent of the differences in school readiness between children born preterm and children born full term in the population-based cohort was not observed in the sibling cohort. This finding could be attributed to unmeasured confounders in the population-based cohort. Children born preterm and their siblings born full term were similar for various sociodemographic factors, whereas children in the population-based cohort were quite different. The E-value that measures unmeasured confounding was 1.95 for a GA less than 34 weeks in the population-based model, with a lower 95% CI of 1.73. The E-value suggests that the observed AOR of 1.72 for less than 34 weeks in the model could only be explained by an unmeasured confounder with an AOR of 1.95 or above, but weaker confounders would not do so. This E-value was higher than the effect estimates of all other covariates in the model except male sex, suggesting the likely presence of unmeasured confounding in the population-based cohort. Furthermore, parental stress associated with a child born preterm and their focus on these children might limit or reduce the quality of their interactions with other children in the family, adversely affecting the development and school performance of the other children.44,45,46 These findings in the sibling cohort are hypothesis generating and need further exploration.
Based on these findings, it becomes relevant to support children born preterm, especially those born extremely preterm, during their early school years. Multiple interventions have been shown to improve school readiness at the population level, such as food supplement programs for women, infants, and children; home visitation by nurses; and primary prevention programs, such as Read Out Loud.47,48,49 For children born preterm specifically, maximizing their exposure to preschool educational programs, early referral to intervention programs, family and community supports to reduce socioeconomic and health disparities and to identify these disparities, and ongoing monitoring at school entry will help to mitigate some of the adverse implications of prematurity for school readiness.50,51 Pediatricians also have a major role in promoting the overall growth and development of these children during the follow-up visits by providing developmental screening, by modeling appropriate parent-child interactions, and by connecting these families to appropriate community resources.
Strengths and Limitations
This study has some strengths. To our knowledge, this school readiness evaluation was the first to use the EDI in a large, population-based cohort of children born preterm while accounting for familial, genetic, and environmental confounders in a sibling cohort. Use of the EDI helped us overcome some of the challenges of existing research, especially studies that limited their assessment to specific domains of school readiness and used nonvalidated tools. It has been shown that restricting school readiness assessment to a specific domain (eg, cognitive domain) is likely to lead to underestimation of later academic achievement.50 The EDI data of children are collected by teachers in the school (natural) environment using observations over the entire school year, unlike other tools that assess these children’s abilities on the basis of a one-time performance in an office or clinic setting. The 2014 data from the Canadian Institute of Child Health showed that 26.7% of children in Canada were vulnerable in the EDI,52 a rate similar to the 28% in the present cohort, suggesting that the sample in this study is representative of the Canadian population.
This study also has some limitations. We had the EDI data for only 75% of the eligible children, partly because of the EDI being administered every other year and partly because of families moving out of Manitoba between the children’s birth and EDI assessment. The population-based cohort also had a relatively small number of children with lower GAs. In addition, we could include only the variables that were available in the data repository.
Conclusions
In this cohort study, children born preterm had a lower rate of school readiness than children born full term in the population-based cohort, but no difference was found in the sibling cohort. There was a GA gradient seen for both the primary and secondary outcomes. In the population-based cohort, various child and maternal factors were associated with vulnerability that can be used to identify and support high-risk children born preterm. In the sibling cohort, the similarities in school readiness between children born preterm and their siblings born full term suggest that some of the poor outcomes for premature neonates may be associated with the social determinants of health and medical factors of prematurity rather than with prematurity itself; these areas warrant further research and interventions.
eTable 1. Definitions of Maternal and Neonatal Variables Used in the Study
eTable 2. Characteristics of All Children Included in the Population-Based Cohort and Those Excluded (Due to Lack of Linkage to the EDI)
eTable 3. EDI Domain Scores for Children Born Preterm and Those Born Full Term in the Population-Based Cohort
eTable 4. EDI Domain Scores for Children Born Preterm and Their Full Term Siblings
eTable 5. Regression Model Results for Vulnerability in EDI Among Population-Based and Sibling Cohorts
eFigure. Violin Plots Showing the Scores in Each Domain of EDI in the Population-Based Cohort
References
- 1.High PC; American Academy of Pediatrics Committee on Early Childhood, Adoption, and Dependent Care and Council on School Health . School readiness. Pediatrics. 2008;121(4):e1008-e1015. doi: 10.1542/peds.2008-0079 [DOI] [PubMed] [Google Scholar]
- 2.Feinstein L, Sabates R, Anderson TM. What are the effects of education on health? In: Desjardins R, Schuller T, eds. Measuring the Effects of Education on Health and Civic Engagement: Proceedings of the Copenhagen Symposium Organisation for Economic Co-operation and Development. UCLA; 2006:171-353. Accessed July 1, 2022. https://www.oecd.org/education/innovation-education/measuringtheeffectsofeducationonhealthandcivicengagement.htm
- 3.Guhn M, Gadermann AM, Almas A, Schonert-Reichl KA, Hertzman C. Associations of teacher-rated social, emotional, and cognitive development in kindergarten to self-reported wellbeing, peer relations, and academic test scores in middle childhood. Early Child Res Q. 2016;35:76-84. doi: 10.1016/j.ecresq.2015.12.027 [DOI] [Google Scholar]
- 4.Jordan NC, Kaplan D, Ramineni C, Locuniak MN. Early math matters: kindergarten number competence and later mathematics outcomes. Dev Psychol. 2009;45(3):850-867. doi: 10.1037/a0014939 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.McMahon WW, Oketch M. Education’s effects on individual life chances and on development: an overview. Br J Educ Stud. 2013;61:79-107. doi: 10.1080/00071005.2012.756170 [DOI] [Google Scholar]
- 6.Aarnoudse-Moens CS, Weisglas-Kuperus N, van Goudoever JB, Oosterlaan J. Meta-analysis of neurobehavioral outcomes in very preterm and/or very low birth weight children. Pediatrics. 2009;124(2):717-728. doi: 10.1542/peds.2008-2816 [DOI] [PubMed] [Google Scholar]
- 7.McBryde M, Fitzallen GC, Liley HG, Taylor HG, Bora S. Academic outcomes of school-aged children born preterm: a systematic review and meta-analysis. JAMA Netw Open. 2020;3(4):e202027. doi: 10.1001/jamanetworkopen.2020.2027 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Twilhaar ES, de Kieviet JF, Aarnoudse-Moens CS, van Elburg RM, Oosterlaan J. Academic performance of children born preterm: a meta-analysis and meta-regression. Arch Dis Child Fetal Neonatal Ed. 2018;103(4):F322-F330. doi: 10.1136/archdischild-2017-312916 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Garfield CF, Karbownik K, Murthy K, et al. Educational performance of children born prematurely. JAMA Pediatr. 2017;171(8):764-770. doi: 10.1001/jamapediatrics.2017.1020 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Patrianakos-Hoobler AI, Msall ME, Marks JD, Huo D, Schreiber MD. Risk factors affecting school readiness in premature infants with respiratory distress syndrome. Pediatrics. 2009;124(1):258-267. doi: 10.1542/peds.2008-1771 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Pritchard VE, Clark CA, Liberty K, Champion PR, Wilson K, Woodward LJ. Early school-based learning difficulties in children born very preterm. Early Hum Dev. 2009;85(4):215-224. doi: 10.1016/j.earlhumdev.2008.10.004 [DOI] [PubMed] [Google Scholar]
- 12.Roberts G, Lim J, Doyle LW, Anderson PJ. High rates of school readiness difficulties at 5 years of age in very preterm infants compared with term controls. J Dev Behav Pediatr. 2011;32(2):117-124. doi: 10.1097/DBP.0b013e318206d5c9 [DOI] [PubMed] [Google Scholar]
- 13.Shah PE, Kaciroti N, Richards B, Lumeng JC. Gestational age and kindergarten school readiness in a national sample of preterm infants. J Pediatr. 2016;178:61-67. doi: 10.1016/j.jpeds.2016.06.062 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Woythaler M, McCormick MC, Mao WY, Smith VC. Late preterm infants and neurodevelopmental outcomes at kindergarten. Pediatrics. 2015;136(3):424-431. doi: 10.1542/peds.2014-4043 [DOI] [PubMed] [Google Scholar]
- 15.Manitoba Centre for Health Policy. The Manitoba Population Research Data Repository—overview. Accessed October 18, 2021. https://umanitoba.ca/faculties/health_sciences/medicine/units/chs/departmental_units/mchp/resources/repository/index.html
- 16.Chateau D, Metge C, Prior H, Soodeen RA. Learning from the census: the Socio-economic Factor Index (SEFI) and health outcomes in Manitoba. Can J Public Health. 2012;103(8 suppl 2):S23-S27. doi: 10.1007/BF03403825 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Nickel NC, Chateau DG, Martens PJ, et al. ; PATHS Equity Team . Data resource profile: Pathways to Health and Social Equity for Children (PATHS Equity for Children). Int J Epidemiol. 2014;43(5):1438-1449. doi: 10.1093/ije/dyu190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Jutte DP, Brownell M, Roos NP, Schippers C, Boyce WT, Syme SL. Rethinking what is important: biologic versus social predictors of childhood health and educational outcomes. Epidemiology. 2010;21(3):314-323. doi: 10.1097/EDE.0b013e3181d61e61 [DOI] [PubMed] [Google Scholar]
- 19.Jutte DP, Roos LL, Brownell MD. Administrative record linkage as a tool for public health research. Annu Rev Public Health. 2011;32:91-108. doi: 10.1146/annurev-publhealth-031210-100700 [DOI] [PubMed] [Google Scholar]
- 20.Irvine B, Luo W, León JA. Congenital anomalies in Canada 2013: a perinatal health surveillance report by the Public Health Agency of Canada’s Canadian Perinatal Surveillance System. Health Promot Chronic Dis Prev Can. 2015;35(1):21-22. doi: 10.24095/hpcdp.35.1.04 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Early Development Instrument . How to interpret EDI results. Accessed September 22, 2021. https://edi.offordcentre.com
- 22.Curtin M, Madden J, Staines A, Perry IJ. Determinants of vulnerability in early childhood development in Ireland: a cross-sectional study. BMJ Open. 2013;3(5):e002387. doi: 10.1136/bmjopen-2012-002387 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Janus M, Offord DR. Development and psychometric properties of the Early Development Instrument (EDI): a measure of children’s school readiness. Can J Behav Sci. 2007;39:1-22. doi: 10.1037/cjbs2007001 [DOI] [Google Scholar]
- 24.Janus M, Brinkman S, Duku E. Validity and psychometric properties of the Early Development Instrument in Canada, Australia, United States, and Jamaica. Soc Indic Res. 2011;103:283-297. doi: 10.1007/s11205-011-9846-1 [DOI] [Google Scholar]
- 25.VanderWeele TJ, Ding P. Sensitivity analysis in observational research: introducing the e-value. Ann Intern Med. 2017;167(4):268-274. doi: 10.7326/M16-2607 [DOI] [PubMed] [Google Scholar]
- 26.Benjamini Y, Krieger AM, Yekutieli D. Adaptive linear step-up procedures that control the false discovery rate. Biometrika. 2006;93(3):491-507. doi: 10.1093/biomet/93.3.491 [DOI] [Google Scholar]
- 27.Anderson PJ, Doyle LW; Victorian Infant Collaborative Study Group . Executive functioning in school-aged children who were born very preterm or with extremely low birth weight in the 1990s. Pediatrics. 2004;114(1):50-57. doi: 10.1542/peds.114.1.50 [DOI] [PubMed] [Google Scholar]
- 28.Bhutta AT, Cleves MA, Casey PH, Cradock MM, Anand KJ. Cognitive and behavioral outcomes of school-aged children who were born preterm: a meta-analysis. JAMA. 2002;288(6):728-737. doi: 10.1001/jama.288.6.728 [DOI] [PubMed] [Google Scholar]
- 29.Botting N, Powls A, Cooke RW, Marlow N. Cognitive and educational outcome of very-low-birthweight children in early adolescence. Dev Med Child Neurol. 1998;40(10):652-660. doi: 10.1111/j.1469-8749.1998.tb12324.x [DOI] [PubMed] [Google Scholar]
- 30.Walther FJ, den Ouden AL, Verloove-Vanhorick SP. Looking back in time: outcome of a national cohort of very preterm infants born in the Netherlands in 1983. Early Hum Dev. 2000;59(3):175-191. doi: 10.1016/S0378-3782(00)00094-3 [DOI] [PubMed] [Google Scholar]
- 31.Walberg HJ, Tsai SL. Matthew effects in education. Am Educ Res J. 1983;20(3):359-373. doi: 10.3102/00028312020003359 [DOI] [Google Scholar]
- 32.National Education Goals Panel . The Goal 1 Technical Planning Subgroup Report on School Readiness. National Education Goals Panel; 1991. [Google Scholar]
- 33.Auger A, Farkas G, Burchinal MR, Duncan GJ, Vandell DL. Preschool center care quality effects on academic achievement: an instrumental variables analysis. Dev Psychol. 2014;50(12):2559-2571. doi: 10.1037/a0037995 [DOI] [PubMed] [Google Scholar]
- 34.Duncan GJ, Dowsett CJ, Claessens A, et al. School readiness and later achievement. Dev Psychol. 2007;43(6):1428-1446. doi: 10.1037/0012-1649.43.6.1428 [DOI] [PubMed] [Google Scholar]
- 35.Pagani LS, Fitzpatrick C. Children’s school readiness: implications for eliminating future disparities in health and education. Health Educ Behav. 2014;41(1):25-33. doi: 10.1177/1090198113478818 [DOI] [PubMed] [Google Scholar]
- 36.D’Angiulli A, Warburton W, Dahinten S, Hertzman C. Population-level associations between preschool vulnerability and grade-four basic skills. PLoS One. 2009;4(11):e7692. doi: 10.1371/journal.pone.0007692 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Murray E, Fernandes M, Fazel M, Kennedy SH, Villar J, Stein A. Differential effect of intrauterine growth restriction on childhood neurodevelopment: a systematic review. BJOG. 2015;122(8):1062-1072. doi: 10.1111/1471-0528.13435 [DOI] [PubMed] [Google Scholar]
- 38.Isaacs EB, Lucas A, Chong WK, et al. Hippocampal volume and everyday memory in children of very low birth weight. Pediatr Res. 2000;47(6):713-720. doi: 10.1203/00006450-200006000-00006 [DOI] [PubMed] [Google Scholar]
- 39.Makhoul IR, Soudack M, Goldstein I, Smolkin T, Tamir A, Sujov P. Sonographic biometry of the frontal lobe in normal and growth-restricted neonates. Pediatr Res. 2004;55(5):877-883. doi: 10.1203/01.PDR.0000119369.21770.7A [DOI] [PubMed] [Google Scholar]
- 40.Tolsa CB, Zimine S, Warfield SK, et al. Early alteration of structural and functional brain development in premature infants born with intrauterine growth restriction. Pediatr Res. 2004;56(1):132-138. doi: 10.1203/01.PDR.0000128983.54614.7E [DOI] [PubMed] [Google Scholar]
- 41.Dahlquist G, Källén B. School marks for Swedish children whose mothers had diabetes during pregnancy: a population-based study. Diabetologia. 2007;50(9):1826-1831. doi: 10.1007/s00125-007-0744-7 [DOI] [PubMed] [Google Scholar]
- 42.Ornoy A, Ratzon N, Greenbaum C, Wolf A, Dulitzky M. School-age children born to diabetic mothers and to mothers with gestational diabetes exhibit a high rate of inattention and fine and gross motor impairment. J Pediatr Endocrinol Metab. 2001;14(suppl 1):681-689. doi: 10.1515/JPEM.2001.14.S1.681 [DOI] [PubMed] [Google Scholar]
- 43.Wall-Wieler E, Roos LL, Gotlib IH. Maternal depression in early childhood and developmental vulnerability at school entry. Pediatrics. 2020;146(3):e20200794. doi: 10.1542/peds.2020-0794 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Elgar FJ, Mills RS, McGrath PJ, Waschbusch DA, Brownridge DA. Maternal and paternal depressive symptoms and child maladjustment: the mediating role of parental behavior. J Abnorm Child Psychol. 2007;35(6):943-955. doi: 10.1007/s10802-007-9145-0 [DOI] [PubMed] [Google Scholar]
- 45.Singer LT, Fulton S, Kirchner HL, et al. Parenting very low birth weight children at school age: maternal stress and coping. J Pediatr. 2007;151(5):463-469. doi: 10.1016/j.jpeds.2007.04.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Vigod SN, Villegas L, Dennis CL, Ross LE. Prevalence and risk factors for postpartum depression among women with preterm and low-birth-weight infants: a systematic review. BJOG. 2010;117(5):540-550. doi: 10.1111/j.1471-0528.2009.02493.x [DOI] [PubMed] [Google Scholar]
- 47.Meyers AF, Sampson AE, Weitzman M, Rogers BL, Kayne H. School breakfast program and school performance. Am J Dis Child. 1989;143(10):1234-1239. doi: 10.1001/archpedi.1989.02150220142035 [DOI] [PubMed] [Google Scholar]
- 48.Mendelsohn AL, Mogilner LN, Dreyer BP, et al. The impact of a clinic-based literacy intervention on language development in inner-city preschool children. Pediatrics. 2001;107(1):130-134. doi: 10.1542/peds.107.1.130 [DOI] [PubMed] [Google Scholar]
- 49.Peacock S, Konrad S, Watson E, Nickel D, Muhajarine N. Effectiveness of home visiting programs on child outcomes: a systematic review. BMC Public Health. 2013;13:17. doi: 10.1186/1471-2458-13-17 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Brooks-Gunn J, McCarton CM, Casey PH, et al. Early intervention in low-birth-weight premature infants: results through age 5 years from the Infant Health and Development Program. JAMA. 1994;272(16):1257-1262. doi: 10.1001/jama.1994.03520160041040 [DOI] [PubMed] [Google Scholar]
- 51.McCormick MC, McCarton C, Tonascia J, Brooks-Gunn J. Early educational intervention for very low birth weight infants: results from the Infant Health and Development Program. J Pediatr. 1993;123(4):527-533. doi: 10.1016/S0022-3476(05)80945-X [DOI] [PubMed] [Google Scholar]
- 52.The Health of Canada’s Children and Youth: A CICH Profile. Figure 8.2.2: percentage of children who are vulnerable in at least one area of development at age 5, by sex, Canada (excluding Nunavut), 2014. Accessed August 22, 2021. https://cichprofile.ca/module/8/section/8/page/percentage-of-children-vulnerable-in-at-least-one-area-of-development-age-5-by-sex-2014
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Definitions of Maternal and Neonatal Variables Used in the Study
eTable 2. Characteristics of All Children Included in the Population-Based Cohort and Those Excluded (Due to Lack of Linkage to the EDI)
eTable 3. EDI Domain Scores for Children Born Preterm and Those Born Full Term in the Population-Based Cohort
eTable 4. EDI Domain Scores for Children Born Preterm and Their Full Term Siblings
eTable 5. Regression Model Results for Vulnerability in EDI Among Population-Based and Sibling Cohorts
eFigure. Violin Plots Showing the Scores in Each Domain of EDI in the Population-Based Cohort