Depression as a Longitudinal Modulator of Physical–Verbal Aggression in Early Adolescence: Trajectories and Sex Comparisons
Description
Background: Depression in early adolescence has been increasingly linked to heightened aggressive behaviour, yet little is known about how depressive symptoms shape the developmental trajectory of physical–verbal aggression during this critical period. Early adolescence (ages 10–14) is marked by the onset of sex differences in depression as well as significant changes in externalising behaviours, making it an essential stage for understanding comorbidity patterns. This study examines whether depressive symptoms function as a chronic and dynamic risk factor for aggression and whether these effects differ by sex. Objective: (1) To analyse the longitudinal effect of depression on physical–verbal aggression across early adolescence; (2) To determine whether sex moderates the developmental trajectory and severity of aggression in adolescents with and without depression. Method: A three‑wave longitudinal design was conducted with 251 adolescents (66.5% girls; ages 10–14) recruited through random probability sampling in Madrid, Spain. Depression was assessed using the CDI‑S, and physical–verbal aggression with the PVA. Repeated‑Measures ANOVA tested intra‑ and inter‑individual differences across ages 10.8 (T1), 11.8 (T2), and 12.8 (T3), including sex and depression status as between‑subjects factors. Results: Depression exerted a significant longitudinal effect on aggression (F(1,249)=14.54, p<.001, η²=.055), with adolescents with depression consistently scoring above global mean thresholds across all three time points. Aggression followed a quadratic trajectory (increase from 10.8 to 11.8 years, subsequent decrease at 12.8 years), but remained at elevated baseline levels when depression was present. Although boys and girls showed similar developmental curves (non‑significant interaction), boys displayed greater overall severity (F(1,249)=6.05, p=.015, η²=.024). Depression had a stronger impact in boys (η²=.073) than in girls (η²=.050), with depressed boys exhibiting the highest and most persistent aggression levels. Conclusions: Depression acts as a chronic and early‑emerging risk factor for physical–verbal aggression during early adolescence. Although trajectories are similar across sexes, boys show greater severity and reduced decline over time, suggesting increased vulnerability. These findings highlight the need for early screening of depressive symptoms in aggressive youth, and underscore the importance of longitudinal assessment for identifying high‑risk profiles and informing sex‑sensitive preventive interventions.
Files
Steps to reproduce
Procedure The selection of the sample was carried out by random probability sampling. Age and sex were taken as population parameters. This study focused on the set of public and private schools provided by the Education Delegation of the Madrid Community. After contacting the director of the centre and commenting on the proposal for data collection proposal, we sent them a summary of the study goals and formally requested their participation in the research, as well as their authorization. If the answer was affirmative, the school was selected for data collection. This same procedure was carried out in 15 schools. Finally, there were 10 who agreed to participate in the study. The planning was carried out jointly by a team of trained psychologists for the task and the teachers in charge of each classroom. The tests were carried out by psychologists in the class groups with a maximum of 25 students per classroom. Before the beginning of the evaluation, the research and the evaluation instruments were briefly explained to the subjects. The Ethics Commission of the National University of Distance Education (UNED) authorised the study and its compliance with the ethical and data protection standards required by European legislation. Statistical analysis A repeated measures ANOVA (RM-ANOVA) was conducted to analyze the longitudinal trajectory of physical and verbal aggression, considering both intraindividual changes and interindividual differences. This statistical method enables the estimation of group mean comparisons for a dependent variable measured across multiple time points (Huck & McLean, 1975). The longitudinal design ensures the appropriate sequencing of variables, facilitating the examination of developmental changes and the inference of potential statistical causal effects, which are not achievable with cross-sectional approaches (Edwards, 2000). RM-ANOVA includes two key analytical factors: within-subject and between-subject. In this study, age served as the within-subject factor, allowing the assessment of changes in physical-verbal aggression over the measurement period, while sex and depression presence-absence functioned as between-subject factors, enabling comparisons across groups (Blanca-Mena et al., 2022). The method also ensures subject homogeneity, improving comparability and controlling for individual differences (Arnau & Bono, 2008; Guardia et al., 2007; León & Montero, 1997). Given its robustness in model construction and statistical power, longitudinal designs have been demonstrated to be more efficient and reliable than cross-sectional designs. The data were analyzed using IBM SPSS Statistics (Version 26).
Institutions
- Universidad a Distancia de MadridCollaro Villalba