Informant Discrepancies and the Heritability of Antisocial Behavior: A Meta-Analysis

TL;DR

This meta-analysis investigates the heritability of antisocial behavior across different informants, ages, and study designs, revealing bimodal heritability distribution and emphasizing the importance of informant perspective in behavioral genetics research.

Contribution

It provides a comprehensive meta-analytic assessment of heritability estimates for antisocial behavior, highlighting the influence of informant type and age, and proposing a flexible model for selecting informants.

Findings

Heritability distribution is bimodal with moderate and high clusters.

Four moderators significantly affect heritability estimates.

Heritability patterns vary across informants and age groups.

Abstract

Antisocial behavior, which includes both aggressive and delinquent activities, is the opposite of prosocial behavior. Researchers have studied the heritability of antisocial behavior among twin and non-twin sibling pairs from behavioral ratings made by parents, teachers, observers, and youth. Through a meta-analysis, we examined longitudinal and cross sectional research in the behavioral genetics of antisocial behavior, consisting of 42 studies, of which 38 were studies of twin pairs, 3 were studies of twins and non-twin siblings, and 1 was a study of adoptees. These studies provided n = 89 heritability (h2) effect size estimates from a total of 94,517 sibling pairs who ranged in age from 1.5 to 18 years; studies provided data for 29 moderators (predictors). We employed a random-effects meta-analysis model to achieve three goals: (a) perform statistical inference of the overall heritability distribution in the underlying population of studies, (b) identify significant study level moderators (predictors) of heritability, and (c) examine how the heritability distribution varied as a function of age and type of informant, particularly in longitudinal research. The meta-analysis indicated a bimodal overall heritability distribution, indicating two clusters of moderate and high heritability values, respectively; identified four moderators that predicted significant changes in mean heritability; and indicated differential patterns of median h2 and variance (interquartile ranges) across informants and ages. We argue for a cross-perspective, cross-setting model for selecting informants in behavioral genetic research, that is flexible and sensitive to changes in antisocial behavior over time.