Recurrent non-suicidal self-injury in depressed youth with mixed features: a 6-month prospective cohort study

Non-suicidal self-injury (NSSI) is highly prevalent among youth and carries substantial clinical significance [1, 2], particularly in individuals with major depressive disorder (MDD) [3, 4]. Epidemiological data indicate that approximately 30% to 60% of youth with MDD report engaging in NSSI [5]. Those who self-injure exhibit markedly elevated risk for suicide and often develop repetitive or chronic self-harming behaviors, leading to significant functional impairment and placing a considerable burden on families and healthcare systems [6, 7].

Mixed features, as defined in the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition (DSM-5), refer to the presence of subthreshold manic symptoms during a depressive episode that do not meet full criteria for mania or hypomania [8‐10]. Approximately 23.8% of individuals with depressive disorders meet criteria for mixed features [11], and this subgroup has been consistently associated with increased risk of suicide attempts [12]. Furthermore, previous studies have indicated that youth with mixed features often exhibit increased emotional arousal and poor impulse control, both of which have been consistently associated with NSSI [13]. Elevated emotional arousal and impaired impulse control may contribute to NSSI by intensifying emotional distress and diminishing inhibitory capacity [14, 15].

Despite these associations, few studies have systematically investigated the impact of mixed features on NSSI, particularly in youth populations. Moreover, there is a paucity of longitudinal data on the developmental trajectories of self-injurious behaviors in this group [16, 17]. To address this gap, the present multicenter prospective study aimed to examine whether the presence of mixed features at baseline predicts the occurrence of NSSI within a six-month follow-up period. Elucidating this potential relationship may contribute to earlier identification of at-risk youth and inform the development of developmentally appropriate, targeted intervention strategies.

Participants were recruited from the South China Adolescent Depression Cohort, a multicenter prospective longitudinal study coordinated by the Affiliated Brain Hospital of Guangzhou Medical University in collaboration with multiple psychiatric institutions across South China. Recruitment took place between March 1, 2022, and March 31, 2024.

Eligible participants were aged 13–25 years, met DSM-5 criteria for major depressive disorder (MDD), and were currently experiencing a major depressive episode (MDE). Individuals were excluded if they had a prior diagnosis of schizophrenia, bipolar disorder, intellectual disability, or other severe psychiatric conditions, or if they exhibited cognitive or language impairments that interfered with assessment completion.

The study protocol was approved by the Ethics Committee of the Affiliated Brain Hospital of Guangzhou Medical University (Approval No. 2022-031), and written informed consent was obtained from all participants. For minors (< 18 years), consent was additionally obtained from a parent or legal guardian.

This study utilized data from a large multicenter cohort in South China to systematically examine the predictive role of mixed features on the risk of recurrent NSSI among youth with depression. Findings demonstrated that mixed features were significantly associated with an increased risk of recurrent NSSI within a six-month follow-up period. This association remained robust across both multiple imputation and complete-case analyses. Stratified analyses further revealed that although mixed features consistently tended to elevate NSSI risk across subgroups, the association was particularly pronounced among individuals experiencing their first depressive episode and those with an illness duration of less than six months.

Although KM survival analysis indicated an earlier onset of NSSI among individuals with mixed features, this association did not reach statistical significance in Cox proportional hazards modeling, suggesting that mixed features may have limited independent prognostic value for incident NSSI. Notably, a prior history of NSSI emerged as the most robust predictor of new-onset NSSI during the follow-up period, underscoring the central importance of behavioral history in risk assessment [28]. In contrast, mixed features were consistently and significantly associated with NSSI recurrence, with hazard ratios ranging from 1.35 to 1.58 across both multiple imputation and complete-case AG models. These findings raise the possibility that mixed features may play a causal role in the persistence and chronicity of self-injurious behaviors.

Our findings expand the current literature on mixed features by shifting the focus from bipolar disorder to depressive episodes. Prior research has primarily examined mixed features in relation to bipolar conversion [29], treatment outcomes [30], and functional impairment [31, 32], with limited attention to their role in NSSI. While some studies have observed elevated NSSI risk among youth with manic or mixed symptoms, such findings have largely been confined to bipolar-spectrum conditions [33]. The present study instead explores mixed features within the context of unipolar depressive episodes, specifically among adolescents and young adults. It is noteworthy that the detection rate of mixed features in our sample was considerably higher than that reported in previous DSM-5–based studies [34]. This discrepancy may be attributable to the composition of the sample, as the majority of participants were adolescents or young adults—a population typically regarded as being at elevated risk for mixed features [35]. With the progression of illness, some individuals within this group may subsequently convert to bipolar disorder, whereas others may evolve toward a relatively stable depressive disorder. By focusing on this population, which has hitherto received limited attention, the present study offers novel insights into the potential association between mixed features in depressive states and self-injurious behavior.

From a biological perspective, mixed features may reflect dysregulation within neurotransmitter systems, including dopamine and serotonin pathways, which are critically involved in affect regulation and impulse control [36, 37]. Psychologically, mixed features are characterized by emotional instability, heightened impulsivity, and behavioral activation, which may impair individuals’ capacity for effective emotion regulation [38, 39]. This dysregulation of emotional control may lead individuals to engage in NSSI as a maladaptive coping strategy aimed at alleviating negative affect. Indeed, emotional dysregulation has been identified as a core mechanism underlying NSSI, particularly among those with pronounced emotional instability [40]. Behavioral theories further suggest that NSSI recurrence is often maintained through negative reinforcement mechanisms [41, 42]. The behavioral activation inherent in mixed features may facilitate the use of NSSI as a means of immediate emotional relief, thereby perpetuating a maladaptive cycle that is resistant to disruption through conventional cognitive control processes [43, 44].

Clinically, these findings underscore the need to expand the identification of mixed features beyond their role in bipolar risk stratification to include sustained monitoring for self-injurious behavior. Notably, the presence of mixed features in depressed adolescents appears to confer heightened vulnerability to persistent, rather than episodic, NSSI. This highlights the importance of targeting the mechanisms—such as heightened emotional reactivity, impulsivity, and behavioral activation—that may maintain self-injurious behaviors over time. Recent studies have increasingly emphasized the need for personalized interventions in individuals with mixed features [45], given that standard treatments for unipolar depression may not adequately address this symptom profile [46]. Mechanism-based approaches, including emotion regulation training, impulsivity management, and behavioral modulation strategies, may be particularly relevant. In light of the chronic nature of NSSI in this subgroup, clinical care should prioritize early identification and sustained, individualized treatment planning.

This study possesses several notable strengths. First, it involves a relatively large sample drawn from a multicenter, prospective cohort, with a specific focus on youth depression, thereby enhancing the clinical relevance and applicability of the findings to this high-risk population. Second, missing data were rigorously addressed through both multiple imputation and complete-case analyses, which improved the robustness and reliability of the results. Third, the study employed multiple statistical models to differentiate risks for both incident and recurrent NSSI, allowing for a more nuanced understanding of the role of mixed features across distinct stages of NSSI risk. Nevertheless, several limitations warrant consideration. First, although diagnoses in this study were based on DSM-5 criteria, the absence of structured clinical interviews such as the SCID may have reduced diagnostic precision. Attention-deficit/hyperactivity disorder (ADHD) was not systematically assessed. Prior research indicates that ADHD increases the likelihood of mixed features among individuals with depression [47]. Undetected or mild ADHD therefore may have contributed to impulsivity or NSSI risk and introduced residual confounding. Other psychiatric comorbidities that are closely related to emotional dysregulation, including borderline personality disorder, eating disorders, and additional neurodevelopmental conditions, were also not formally evaluated because they were not included in the original study protocol. Although participants with severe impairments that prevented completion of the assessment were excluded, milder or undiagnosed comorbid conditions may have remained unrecognized. As a result, some residual confounding related to emotional dysregulation or unmeasured psychiatric conditions cannot be excluded. Second, childhood trauma and emotional dysregulation were not assessed in this cohort. Both constructs are strongly associated with NSSI, and the absence of these measures may have contributed additional residual confounding. Third, the six-month follow-up period may not be sufficient to capture the longer-term course of NSSI. Fourth, reliance on self-reported data for self-injurious behaviors introduces the possibility of recall bias. Fifth, several potentially important covariates, including medication use, psychological interventions, and family environment, were not assessed, which may limit the ability to fully account for factors influencing NSSI risk. Finally, the sample was recruited exclusively from South China. This geographic restriction may limit the generalizability of the findings to other regions or cultural contexts. Despite these limitations, the study provides novel longitudinal evidence regarding the contribution of mixed features to the persistence of self-injury in adolescents and young adults with depression.

In conclusion, while mixed features show limited predictive value for the initial onset of NSSI during follow-up, they demonstrate significant and independent prognostic relevance for NSSI recurrence. Importantly, this effect was consistent across different subgroups. These findings highlight the critical need to integrate mixed features into risk assessment and intervention strategies for self-injurious behaviors in adolescents and young adults with depression, especially among those exhibiting emotional instability and impulsivity. Future research should aim to extend follow-up periods, incorporate a wider array of clinical and environmental covariates, and elucidate the mediating mechanisms linking mixed features to NSSI, ultimately informing the development of personalized, mechanism-driven interventions.