Fatigue emerges as key depression target in trauma-exposed adolescents

Childhood abuse and childhood neglect are often lumped together under the single umbrella of early adversity, but a new study of more than 6,000 trauma-exposed adolescents suggests they may shape depression in partially different ways. Using network analysis, researchers mapped how the nine symptoms of depression connect to one another in young people with different trauma histories — and found that the architecture of the symptom web, not just its severity, differs depending on whether a child was abused, neglected, or both. The findings, published in BMC Psychology, point to fatigue as a promising treatment target across all trauma profiles, while flagging trauma-specific symptoms for prevention.

The research team, led by Zheyu An and Yanqiang Tao of East China Normal University’s Shanghai Key Laboratory of Mental Health and Psychological Crisis Intervention, recruited 6,094 adolescents who had reported childhood trauma. Based on cutoff scores on the Childhood Trauma Questionnaire–Short Form, participants were sorted into three groups: 437 who experienced abuse only, 3,489 who experienced neglect only, and 2,168 who experienced both forms of adversity combined. Depressive symptoms were measured with the Patient Health Questionnaire-9, the standard clinical screener covering sad mood, anhedonia, sleep problems, fatigue, appetite changes, guilt, concentration difficulties, psychomotor changes, and suicidal ideation.

Network analysis treats depression not as a single underlying entity but as a system of symptoms that activate and reinforce one another. In this framework, each symptom is a node, and statistically significant associations between symptoms are edges. The researchers estimated Ising network models, a class of statistical models well suited to binary symptom data, using regularization methods — the extended Bayesian information criterion and enhanced least absolute shrinkage and selection operator — to keep the networks sparse and interpretable. They then compared the three networks on structure, global strength, and centrality using network comparison tests.

The results revealed a striking common core. Fatigue, anhedonia, and sleep disturbances were the most prevalent symptoms in every trauma group, suggesting that regardless of the type of early adversity, the physical and motivational faces of depression dominate. Severity, however, climbed with cumulative exposure: adolescents in the combined-trauma group reported the most intense depressive symptoms, consistent with the idea that abuse and neglect act additively rather than interchangeably.

Centrality — a measure of how strongly a symptom connects to the rest of the network — differed meaningfully across groups. In the abuse-only network, psychomotor changes and concentration difficulties were the most central symptoms, even though the authors caution that centrality stability was low in this small subgroup, with a correlation stability coefficient of 0.128. In the neglect-only network, fatigue and sad mood occupied the hub positions, while in the combined-trauma network, fatigue was the most central symptom. These hub symptoms are theoretically important because they are the points where interventions might propagate most efficiently through the network.

Global strength comparisons added a second layer of differentiation. The abuse-only network was significantly more strongly connected than the neglect-only network, with a test statistic of 7.81 and a p-value below 0.001, and also more strongly connected than the combined-trauma network, with a statistic of 6.89 and p below 0.001. Network structure, by contrast, differed only between the abuse-only and combined-trauma groups, with a statistic of 1.32 and p equal to 0.03. In plain terms, abuse appears to knit depressive symptoms into a tighter, more densely reinforcing web than neglect does — a pattern that could help explain why abused adolescents may be at particular risk once depression takes hold.

To move from descriptive mapping to actionable targets, the researchers ran computer-simulated interventions using the NodeIdentifyR algorithm. The approach works by computationally alleviating or aggravating a single node and measuring how much the rest of the network changes — a digital stress test for candidate symptoms. Under simulated alleviation, fatigue emerged as the standout candidate across all three trauma profiles, with change rates of 31.1 percent in the abuse-only group, 58.8 percent in the neglect-only group, and 42.9 percent in the combined-trauma group. Anhedonia — the loss of pleasure and interest — was identified as a second effective alleviation target in the combined-trauma group, with a change rate of 43.1 percent.

The simulated aggravation runs, which identify symptoms whose worsening would most destabilize the network, flagged different prevention targets for each profile. Concentration difficulties were the top candidate in the abuse-only group, with a change rate of 36.1 percent. Psychomotor changes — observable slowing or agitation — dominated the neglect-only network with a change rate of 92.7 percent, the largest figure in the study. Suicidal ideation was the leading prevention target in the combined-trauma group, with a change rate of 34.6 percent, underscoring the clinical urgency of this highest-severity subgroup.

The authors conclude that abuse-related and neglect-related childhood adversities may be associated with partially distinct depressive symptom pathways, and that trauma-informed interventions may benefit from targeting fatigue across trauma profiles while tailoring prevention strategies to trauma-specific symptom patterns. For clinicians, the message is twofold: a shared treatment entry point exists, but the symptoms most worth monitoring to prevent escalation differ by trauma history.

As with any network study, important caveats apply. The cross-sectional design cannot establish causal direction, and the low centrality stability in the abuse-only subgroup means its hub symptoms should be interpreted with caution. Still, the scale of the sample and the convergence of prevalence, centrality, and simulation results around fatigue give the finding unusual weight. If future longitudinal and intervention studies confirm that alleviating fatigue efficiently deactivates depressive networks in trauma-exposed adolescents, this common symptom — often dismissed as mere tiredness — could become a focal point for trauma-informed mental health care.

Subject of Research: Abuse, neglect, and adolescent depression: distinct symptom networks and potential intervention targets

Article Title: Abuse, neglect, and adolescent depression: distinct symptom networks and potential intervention targets

Article References: An, Z., Chen, R., Lu, X., Wang, Y., Yang, H., Liu, X., Zhou, J., Yan, C., & Tao, Y. (2026). Abuse, neglect, and adolescent depression: distinct symptom networks and potential intervention targets. BMC Psychology. https://doi.org/10.1186/s40359-026-05603-1

Image Credits: AI Generated

DOI: 10.1186/s40359-026-05603-1

Keywords: Abuse, neglect, adolescent, depression, distinct, symptom, networks, potential, intervention, targets, scientific research