Birth Order Differences in First-Year Neurodevelopment

TL;DR

This study explores if firstborn and second-born children show differences in brain development during their first year of life.

Contribution

The study provides empirical evidence on birth order effects on early neurodevelopment.

Findings

Firstborn children showed slightly higher neurodevelopmental scores compared to second-born children.

The differences were most noticeable in motor and social development domains.

Results suggest potential environmental or parental factors influencing early development.

Abstract

This cohort study examines whether firstborn and second-born children differ in neurodevelopmental outcomes during infancy.

Tables

Table 1.. Estimates From a Mother Fixed-Effects Model of Ages & Stages Questionnaires, Third Edition, Domain Scores by Birth Order Among 2117 Sibling Dyads

Domain	Age 6 mo		Age 12 mo
	EMM (SE)a	Mean difference, β (95% CI)b	EMM (SE)a	Mean difference, β (95% CI)b
Communication
Firstborns	47.7 (0.6)	0 [Reference]	38.4 (0.8)	0 [Reference]
Second-borns	45.8 (0.5)	−1.9 (−3.7 to −0.1)c	36.5 (0.8)	−1.9 (−4.3 to 0.6)
Gross motor
Firstborns	35.9 (0.8)	0 [Reference]	44.7 (1.2)	0 [Reference]
Second-borns	29.5 (0.7)	−6.4 (−8.9 to −4.0)c	41.2 (1.1)	−3.6 (−7.1 to 0.02)
Fine motor
Firstborns	43.2 (0.9)	0 [Reference]	49.8 (0.9)	0 [Reference]
Second-borns	37.2 (0.9)	−5.9 (−8.8 to −3.1)c	45.9 (0.8)	−3.9 (−6.8 to −1.1)c
Problem-solving
Firstborns	46.7 (0.9)	0 [Reference]	41.5 (0.9)	0 [Reference]
Second-borns	40.1 (0.8)	−6.7 (−9.3 to −4.0)c	43.3 (0.8)	1.7 (−1.1 to 4.6)
Personal-social
Firstborns	39.4 (0.8)	0 [Reference]	39.5 (0.9)	0 [Reference]
Second-borns	25.6 (0.9)	−13.8 (−16.7 to −10.9)c	34.8 (0.9)	−4.8 (−7.7 to −1.9)c

Table 2.. Mother Fixed-Effects Estimates of Birth Order Differences in Parental Engagement

Variablea	Mean difference, β (95% CI)b
Birth order, second-born vs firstborn	−0.77 (−1.25 to −0.29)c
Sex, female vs male	0.01 (−0.13 to 0.16)
Birth year
2012 vs 2011	−0.30 (−0.54 to −0.06)c
2013 vs 2011	−0.30 (−0.76 to 0.16)
2014 vs 2011	−0.25 (−0.83 to 0.34)

Full text

Introduction

Birth order differences in cognitive and developmental outcomes are well documented, with firstborn children typically outperforming their later-born siblings.^1^ Several explanations have been proposed, including parental learning and changes in family circumstances. In infancy, when parental engagement plays a central role, resource dilution has been suggested as a contributing mechanism.^2,3^ However, only a few studies have compared siblings within the same family, particularly in the first year of life, to separate birth-order factors from stable family factors. We, therefore, examined whether firstborn and second-born children differ in neurodevelopmental outcomes during infancy using a mother fixed-effects design.

Methods

This cohort study used a within-family design to compare pairs of firstborn and second-born singleton children enrolled in the Japan Environment and Children’s Study (JECS).^4^ Ethical approval was obtained from relevant institutional review boards, and written informed consent was collected. This study was conducted in accordance with STROBE reporting guidelines.

Participants who became pregnant twice within the JECS recruitment period could re-enroll, enabling identification of sibling pairs. The key birth characteristics of JECS participants, such as sex ratio, preterm birth rate, and low birth weight rate, closely resemble those of the Japanese Vital Statistics.^5^

Neurodevelopment was evaluated using the Ages & Stages Questionnaires, Third Edition (ASQ-3) at 6 and 12 months. Participants were compensated with gift certificates (valued at ¥1000 for each questionnaire and ¥3000 for biological sample donation). Parental engagement was measured using a 5-item composite score (range, 0-18, with higher scores indicating greater engagement) capturing the frequency of parent-child play, outdoor activities, and reading aloud (eAppendix and eFigure in Supplement 1). Birth order differences were estimated using mother fixed-effects linear models, comparing siblings within the same family and removing all shared, time-invariant family characteristics. Models included child sex and birth year indicators. SEs were clustered at the mother level. Parental engagement was analyzed using analogous fixed-effects models. Given the conceptual limitations of causal mediation within fixed-effects designs, we did not conduct formal mediation analysis; instead, parental engagement differences are presented descriptively within the same framework. Analyses were conducted using SAS statistical software version 9.4 (SAS Institute) and R statistical software version 4.5.1 (R Project for Statistical Computing). Statistical significance was defined as 2-sided P < .05. Estimates were derived from fixed-effects linear models, and 95% CIs were used to assess statistical significance.

Results

A total of 2117 sibling dyads were included. Estimated marginal means indicated that second-born children scored lower than firstborns across multiple ASQ-3 domains at both 6 and 12 months (Table 1). At 6 months, differences ranged from 1.9 points in communication to 13.8 points in personal-social development. At 12 months, differences persisted for fine motor and personal-social domains, whereas estimates for communication and problem-solving were small and not statistically significant. Sensitivity analyses adjusting for birth weight yielded nearly identical results (changes ≤0.1 points).

Parental engagement scores were also lower for second-born children (Table 2). Fixed-effects models indicated a mean difference of −0.77 points (95% CI, −1.26 to −0.29 points) after adjustment for sex and birth year indicators.

Discussion

Birth order differences favoring firstborn children are well established,^1,2,3^ and the findings of this cohort study extend this evidence by demonstrating that such differences appear during the first year of life. Parallel reductions in parental engagement for second-born infants are consistent with the resource dilution model, suggesting that reduced individualized parental time may contribute to early developmental variation. Some differences attenuated by 12 months, which may be partly explained by the confluence model^6^: later-born infants may benefit from observing and interacting with older siblings, compensating for reduced parental resources.

Limitations include that ASQ-3 scores are parent reported and may reflect birth order–related differences in parental perception. Although all ASQ-3 domains are scored on the same scale, differences varied across domains at 6 months, with the personal-social domain showing the largest gap. These differences narrowed by 12 months, and their long-term clinical meaning remains uncertain. In conclusion, this within-family analysis shows that small but consistent birth order differences in neurodevelopment are already observable during the first year of life.