IMPORTANCEAlthough infancy is the most rapid period of postnatal growth and development, factors associated with variation in infant traits are not well understood. OBJECTIVE To synthesize the large twin study literature partitioning phenotypic variance in psychological traits and developmental milestones in infancy into estimates of heritability and shared and nonshared environment. DATA SOURCES PubMed, PsycINFO, and references of included publications were searched up to February 11, 2021. STUDY SELECTION Peer-reviewed publications using the classical twin design to study psychological traits and developmental milestones from birth to 2 years old were included. DATA EXTRACTION AND SYNTHESIS Data were extracted in line with Preferred Reporting Items for Systematic Reviews and Meta-Analyses and categorized using the International Classification of Functioning, Disability and Health: Children and Youth Version. Data were pooled in 3-level random effects models, incorporating within-cohort variance in outcome measurement and between-cohort variance. Data were analyzed from March 2021 through September 2021.
MAIN OUTCOMES AND MEASURESThe primary outcomes were monozygotic and dizygotic twin correlations. These were used to calculate genetic and shared and nonshared environment estimates.
RESULTSAmong 139 publications that were systematically retrieved, data were available on 79 044 twin pairs (31 053 monozygotic and 47 991 dizygotic pairs), 52 independent samples, and 21 countries. Meta-analyses were conducted on psychological traits and developmental milestones from 106 publications organized into 10 categories of functioning, disability, and health. Moderate to high genetic estimates for 8 categories were found, the highest of which was psychomotor functions (pooled h 2 , 0.59; 95% CI, 0.25-0.79; P < .001). Several categories of traits had substantial shared environment estimates, the highest being mental functions of language (pooled c 2 , 0.59; 95% CI, 0.24-0.86; P = .001). All examined categories of traits had moderate or high nonshared environment estimates, the highest of which were emotional functions (pooled e 2 , 0.42; 95% CI, 0.33-0.50; P < .001) and family relationships (pooled e 2 , 0.42; 95% CI, 0.30-0.55; P < .001).CONCLUSIONS AND RELEVANCE These findings may be an important source of information to guide future gene discovery research, public perspectives on nature and nurture, and clinical insights into the degree to which family history and environments may estimate major domains of infant functioning, disability, and health in psychological traits and developmental milestones.