Thought problems from adolescence to adulthood: measurement invariance and longitudinal heritability

Abstract

This study investigates the longitudinal heritability in Thought Problems (TP) as measured with ten items from the Adult Self Report (ASR). There were ~9,000 twins, ~2,000 siblings and ~3,000 additional family members who participated in the study and who are registered at the Netherlands Twin Register. First an exploratory factor analysis was conducted to examine the underlying factor structure of the TP-scale. Then the TP-scale was tested for measurement invariance (MI) across age and sex. Next, genetic and environmental influences were modeled on the longitudinal development of TP across three age groups (12-18, 19-27 and 28-59 year olds) based on the twin and sibling relationships in the data. An exploratory factor analysis yielded a one-factor solution, and MI analyses indicated that the same TP-construct is assessed across age and sex. Two additive genetic components influenced TP across age: the first influencing TP throughout all age groups, while the second arises during young adulthood and stays significant throughout adulthood. The additive genetic components explained 37% of the variation across all age groups. The remaining variance (63%) was explained by unique environmental influences. The longitudinal phenotypic correlation between these age groups was entirely explained by the additive genetic components. We conclude that the TP-scale measures a single underlying construct across sex and different ages. These symptoms are significantly influenced by additive genetic factors from adolescence to late adulthood.

Fig. 1

Path diagram for longitudinal ADE model on thought problems (TP) for three age groups. The figure shows data from an opposite sex twin pair (TW1 = male, TW2 = female) and their two siblings (BR, brother, SIS, sister). The rectangles represent the log-transformed TP sum-scores (TP1 = TP measured at ages 12–18, TP2 = 19–27, TP3 = 28–59). The circles the latent unmeasured factors (A, additive genetic effects; D, dominant genetic effects; E, non-shared environmental effects, and is omitted in the figure for simplicity, but is modeled in a similar way). In parameter subscripts, m stands for male and f stands for female