Citation

Fernandez Castelao C, Kröner-Herwig B. J. Youth Adolesc. 2014; 43(5): 775-789.

Affiliation

Department of Clinical Psychology and Psychotherapy, Georg-August-University of Göttingen, Goßlerstraße 14, 37073, Göttingen, Germany, c.fernandezcastelao@psych.uni-goettingen.de.

Copyright

(Copyright © 2014, Holtzbrinck Springer Nature Publishing Group)

DOI

10.1007/s10964-013-0011-9

PMID

24002677

Abstract

Recent studies have shown that the development of externalizing behavior in childhood and adolescence can be described through different developmental pathways. However, knowledge about differences between the sexes regarding the trajectories is limited. This study focused on potential differences by examining the trajectories of self-reported externalizing symptoms for girls and boys separately. In addition, the relationships of several familiar and child-specific variables with those developmental courses were assessed. The study was conducted on a large community sample of German youths (N = 3,893; mean age 11.38 years; 50 % girls) over 4 years. Using growth mixture modeling, three different classes of trajectories were found for both sexes. The classes differed with regard to the level and the course of symptoms ("low", "moderate", "high-decreasing"). Girls were overrepresented in the "low" class, whereas boys were predominant in the "moderate" and "high-decreasing" classes. The multiple group analysis revealed that the girls and boys differed significantly in their level and linear course of symptoms with regard to the "high-decreasing" class. In contrast, no sex differences were found in the growth factors of the "low" and "moderate" classes. The regression analyses showed that the children's depressive symptoms, dysfunctional parenting style, and negative family climate were associated significantly with the level and course of symptoms as well as the class membership of girls and boys. Life events predicted class membership only for boys, whereas maternal depressive symptoms and family conflict did not demonstrate any significant relationship. The sizes of the predictive associations with the growth factors were similar for both sexes. The results are discussed with regard to existing developmental models and their possible implications for prevention and future research.

Language: en