Latent patterns of rumination and hopelessness on self-harm behaviors in adolescence.

Latent patterns of rumination and hopelessness on self-harm behaviors in adolescence - PMC Skip to main content An official website of the United States government Here's how you know Here's how you know Search Log in Primary site navigation Search Log in Search PMC Full-Text Archive Search in PMC Journal List User Guide As a library, NLM provides access to scientific literature. Inclusion in an NLM database does not imply endorsement of, or agreement with, the contents by NLM or the National Institutes of Health. Learn more: PMC Disclaimer | PMC Copyright Notice Child Adolesc Psychiatry Ment Health . 2026 Jan 29;20:77. doi: 10.1186/s13034-026-01029-0 Latent patterns of rumination and hopelessness on self-harm behaviors in adolescence Mengting Wang 1, 2, 4 , Shuwen Chen 1, 2, 4 , Jingyi Gao 1, 2, 4 , Cheng Cheng 2, 3, 4 , Li Yang 2, 3, 4 , Hui Ai 2, 3, 4, 5, ✉ Author information Article notes Copyright and License information 1 School of Education, Tianjin University, Tianjin, China 2 Institute of Applied Psychology, Tianjin University, Tianjin, China 3 Academy of Medical Engineering and Translational Medicine, Tianjin University, Tianjin, China 4 Laboratory of Suicidal Behavior Research, Tianjin University, Tianjin, China 5 School of Psychology, Shenzhen Key Lab of Affective and Social Cognitive, Shenzhen University, Shenzhen, China ✉ Corresponding author. Received 2025 Aug 6; Accepted 2026 Jan 5; Collection date 2026. © The Author(s) 2026 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data. PMC Copyright notice PMCID: PMC13231778  PMID: 41612475 Abstract Objectives Self-harm behaviors among adolescents, including suicide attempts (SA) and non-suicidal self-injury (NSSI), present critical public health challenges globally. While past-oriented rumination and future-oriented hopelessness are established core dimensions of maladaptive self-cognition, their distinct associations with different forms of self-harm remain poorly characterized. Methods Using latent profile analysis, we investigated the heterogeneous profiles of rumination and hopelessness in a sample of 951 adolescents (M age : 16.58; male: 420). We further examined how these latent profiles differ in their associations with SA and NSSI, controlling for general affective symptoms. Results Three distinct cognitive profiles were identified: high rumination–high hopelessness ( n = 77, 8.09%), moderate rumination–moderate hopelessness ( n = 531, 55.84%), and low rumination–low hopelessness ( n = 343, 36.07%). Specifically, the high rumination–high hopelessness profile emerged as being uniquely associated with SA, even after controlling for affective symptoms. NSSI was related to the severity of depressive symptoms, but not related to the specific cognitive profiles. Conclusion Our study provides novel insights that the interaction between rumination and hopelessness generates distinct cognitive phenotypes, which show different associations with SA and NSSI. These findings address a key theoretical gap in self-harm mechanisms and suggest the need to reshape prevention paradigms by enabling phenotype-specific interventions targeting cognitive constrictions for at-risk youth. Supplementary Information The online version contains supplementary material available at 10.1186/s13034-026-01029-0. Keywords: Rumination, Hopelessness, Self-harm behaviors, Latent profile analysis Introduction Self-harm behaviors comprise a spectrum of injurious behaviors directed at oneself, ranging from non-suicidal self-injury (NSSI), which is not intended to result in death, to suicide attempts (SA), which involve clear suicidal intent, and ultimately to those who died by suicide [ 1 ]. Globally, more than 720,000 individuals die by suicide each year [ 2 ], making it the second leading cause of death among adolescents and young adults [ 3 ]. Recent epidemiological research indicates that NSSI is highly prevalent, with an estimate lifetime prevalence of approximately 22.0% among adolescents worldwide, and a recurrence rate of 20.3% [ 4 ]. The burden is particularly pronounced in China, where the lifetime prevalence is even higher, estimated at 24.7% among youths [ 5 ]. With detection rates of 22.8% for NSSI and 2.9% for SA among Chinese high school students [ 6 ], these behaviors constitute major public health concerns among Chinese adolescents. Clarifying the key factors associated with SA and NSSI in adolescents may inform the development of early intervention and targeted prevention strategies. Adolescence is a critical period for the formation of self-identity. During this developmental stage, individuals integrate past, present, and future self-related experiences to construct a coherent sense of self, which significantly influences psychological well-being [ 7 ]. Self-harm behaviors are understood as a dynamic process influenced by temporal factors [ 8 ]. A recent study has identified that, among various predictors of suicide behaviors, hopelessness regarding the future constituted the strongest risk factor [ 9 ]. Furthermore, rumination on past events has also been identified as an important transdiagnostic predictor of mental disorders [ 10 ]. Therefore, incorporating both past-oriented and future-oriented psychological constructs may enhance our understanding of the cognitive profiles associated with adolescent self-harm behaviors. Moreover, while the development of SA and NSSI in adolescents may co-occur and share certain affective and behavioral features, they are theoretically differentiated by the presence of suicidal intent, the temporal orientation of the underlying motivation (e.g., future-oriented hopelessness in SA vs. immediate affect regulation in NSSI), and the development of acquired capability for suicide [ 11 – 13 ]. Distinguishing these two different types of self-harm behaviors based on their respective self-cognitive constructs is crucial for informing precise and individualized interventions. Hopelessness is a cognitive construct closely associated with future-oriented thinking. The core components of hopelessness include the abandonment of goals and aspirations, as well as negative expectations about the future [ 14 ]. Hopelessness has been found to be associated with both suicide behaviors and NSSI, as individuals with higher levels of hopelessness exhibited an increased propensity for NSSI [ 15 , 16 ]. Rumination is a maladaptive form of cognitive processing and an ineffective emotion regulation strategy [ 17 ]. Characterized by repetitive focus on past events and the causes and consequences of negative emotions, meta-analytic evidence suggests that rumination predicts both short-term and life-time SA among individuals with depressive disorders as well as in non-clinical populations [ 18 ]. Furthermore, studies have found rumination to be positively associated with the frequency of NSSI [ 19 ]. Individuals exhibiting higher levels of rumination tend to engage in more frequent NSSI when experiencing depressive symptoms [ 20 ]. A previous study has found that rumination moderated the mediating role of hopelessness in the relationship between self-criticism and NSSI [ 21 ]. Nonetheless, no study has jointly examined the roles of past-oriented rumination and future-oriented hopelessness in both SA and NSSI. Hopelessness and rumination have also been identified to have transdiagnostic predictive effects on anxiety and depression [ 22 , 23 ]. However, it remains unclear whether the association between combined hopelessness-rumination and self-harm behaviors were independent of depressive and anxiety symptoms. In summary, the formation of self-identity in adolescents is associated with cognitive processing directed toward both the past and the future. Our study aims to identify the core features of adolescents’ future-oriented hopelessness and past-oriented rumination, and to examine how these features are associated with different self-harm behaviors. We hypothesized that adolescents will exhibit distinct latent profiles characterized by different combinations of rumination and hopelessness levels, and these profiles will show differential associations with SA and NSSI, reflecting potentially distinct risk pathways. To this end, we will use latent profile analysis to identify the potential categories of the combined traits of rumination and hopelessness in adolescents. Then, based on these categories, we will use hierarchical logistic regression to analyze the relationship of these combined traits with different self-harm behaviors. To isolate the specific contribution of latent profiles of rumination and hopelessness beyond affective symptoms, all key associations will be statistically adjusted for depressive and anxiety symptom severity. Method Participants Participants were selected from first-grade and second-grade students at a high school in western China. After screening for response quality (inconsistency between positive/negative questions and responses, N = 40), 951 valid questionnaires were obtained, accounting for 95.96% of the total. Of those, 420 (44.16%) were male and 531 (55.84%) were female. This study received ethical approval from the University Ethical Review Board. Informed consent was obtained from the participants and their primary caregivers prior to conducting the study. Consistent with institutional and school policies, any participants who endorsed high-risk items during the screening process were referred to the school for appropriate follow-up and psychological support. Measures Ruminative responses scale (RRS) Rumination was assessed using the revised Chinese version of the Ruminative Responses Scale (RRS), a 22-item self-report measure assessed on a 4-point Likert scale. This scale comprises three subscales: brooding, depressive symptoms, and reflection. This adaptation has been previously administered in samples of Chinese high school and undergraduate students, demonstrating established reliability and validity within this socio-cultural context [ 24 , 25 ]. To minimize the issue of multicollinearity arising from the significant variance shared by the overall RRS and measures of affective symptoms, only the Brooding and Reflection subscales were selected as indicators for the LPA. This selection aligns with the empirically supported distinction that these two subscales represent the core, maladaptive and adaptive cognitive components of rumination, respectively, as validated by Treynor et al. [ 26 ]. In our sample, the Cronbach’s alpha coefficient for the scale was 0.96. Beck Hopeless Scale (BHS) Hopelessness, defined as an individual’s negative expectations regarding the future, was measured using the Chinese version of the Beck Hopelessness Scale (BHS) [ 27 ]. This adaptation has demonstrated strong psychometric properties in large samples of Chinese adolescents, with a previously reported internal consistency of α = 0.85, supporting its suitability for use with Chinese youth. The scale has 20 entries and measures three dimensions: feelings about the future, loss of motivation, and future expectations. Total scores range from 0 to 20, with higher scores indicating higher levels of hopelessness. Since this study examined cognitive processing in the temporal dimension, the feelings about the future and future expectations were used as observational variables in the latent profile analysis because they were closely related to the future. In this study’s sample, Cronbach’s alpha coefficient was 0.72. State-Trait Anxiety Inventory (STAI) Anxiety symptoms were assessed using the Chinese version of State-Trait Anxiety Inventory (STAI), which has been validated for secondary school students in China [ 28 ]. The first 20 items assess state anxiety (S-AI), reflecting individuals’ transient experiences of stress and worry at a given moment, while the remaining 20 items assess trait anxiety (T-AI), representing a relatively stable predisposition to experience anxiety. Each item is rated on a 4-point scale, ranging from 1 (“not at all”) to 4 (“very much so”), yielding total scores between 20 and 80. Higher scores indicate greater levels of state anxiety. In our study, only the state anxiety subscale was used to assess adolescents’ current anxiety levels. The Cronbach’s alpha coefficient for this subscale was 0.87. Beck Depression Inventory-II (BDI-II) Depressive symptoms over the past week were assessed using the Chinese version of the Beck Depression Inventory-II (BDI-II). This scale has demonstrated high reliability when administered in samples of Chinese middle-school students [ 29 ]. The questionnaire comprises 21 items, each rated on a 4-point scale ranging from 0 to 3. The total score, calculated by summing all item responses, ranges from 0 to 63. Based on established cut-off scores, a total score of 0–13 indicates minimal or no depression, 14–19 indicates mild depression, 20–28 indicates moderate depression, and 29–63 indicates severe depression. In the current sample, the Cronbach’s alpha coefficient for the scale was 0.91. Self-Harm Behaviors Assessment Non-suicidal self-injury (NSSI) and suicide attempts (SA) were assessed through explicit questions. NSSI was evaluated by asking participants: “Have you ever intentionally hurt yourself without the intent to die?” Similarly, SA was assessed with the question: “Have you ever attempted suicide or engaged in any behavior that could be considered suicidal?”. Both SA and NSSI were measured as life-time behaviors. Single-item assessment has been shown to be a reasonable method for capturing lifetime SA/NSSI [ 30 ]. Data analysis Demographics data were analyzed by using SPSS version 26.0. Latent Profile Analysis (LPA), robust three-step approach (R3STEP) and Bolck, Croon, and Hagenaars’s method (BCH) analyses were conducted in Mplus version 8.3. Subsequent hierarchical logistic regression based on LPA derived categories were performed using R version 4.4.1. LPA was used to explore joint patterns of adolescent rumination (i.e., brooding and reflection subscales of the RRS) and hopelessness about the future (i.e., feelings about the future and future expectations from BHS). Prior to conducting LPA, all variables were standardized using z-score transformation. We used the MLR estimator, with 200 random starts and 50 final optimizations to minimize the risk of local maxima. The model was estimated using the default Mplus random seed and default convergence criteria, and all solutions converged normally without indications of local maxima. Model selection followed the criteria [ 31 ]: (1) lower values of the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), and sample-size adjusted BIC (SSA-BIC) indicated better model fit; (2) a significant p -value for the Lo–Mendell–Rubin Test (LMR) and the Bootstrap Likelihood Ratio Test (BLRT) suggested that the model with k profiles fits better than the k–1 model; (3) higher entropy values (closer to 1) indicated greater classification accuracy. Additionally, average posterior probabilities, the smallest class size, model parsimony, and interpretability were considered in determining the optimal solution. In subsequent analyses, the latent class variable was dummy-coded and included as an independent variable. Then, R3STEP was used to analyze the covariates, and the BCH method was applied to the analysis of SA and NSSI [ 31 , 32 ]. To isolate the associations of the rumination–hopelessness profiles after controlling for covariates, hierarchical logistic regression was conducted to further examine their effects on SA and NSSI. Variables were entered in three steps using a general linear model (GLM) framework. In Model 1, demographic variables (gender, age, and grade) were included as covariates. Model 2 added affective disorders (depression and state anxiety) to Model 1. Model 3 further included rumination–hopelessness latent profiles. Categorical variables were converted into factors with designated reference categories. For each model, regression coefficients (β), standard errors (SE), p -values, and odds ratios (ORs) were reported. Model fit was compared across the three nested models using the Likelihood Ratio Test (LRT). A significant difference in deviance ( p < 0.05) indicated that the added variables provided incremental explanatory power. Results Common method biases Harman’s one-factor test was used. The results showed that the total number of factors with eigenvalues greater than 1 was 12, and the total variance explained by the first common factor was 24.76%. Therefore, there was no serious common method bias in the measurement data of this study. Demographic differences and descriptive statistics The life-time prevalence of SA was 12.41% (95% CI 10.30%–14.51%), and the life-time prevalence rate of NSSI was 7.26% (95% CI 5.61%–8.91%). Given the relatively low rate of self-harm behaviors, we included a brief precision assessment in our analytic strategy. To ensure transparency regarding precision and the expected uncertainty, all key estimates from the R3STEP analysis, BCH comparisons, and hierarchical logistic regression models were reported with 95% confidence intervals. Descriptive statistics and bivariate correlations are presented in Table 1 . The results indicated that both rumination and hopelessness were significantly correlated with anxiety, depression, SA and NSSI. In addition, demographic variables, including age, grade, and gender, were also associated with SA. Grade and gender were associated with NSSI. Therefore, these demographic variables were included as covariates in the subsequent analyses of variance. Table 1. Descriptive statistics and bivariate correlations 1 2 3 4 5 6 7 8 9 10 1.Age 1 2.Grade 0.29** 1 3.Gender 0.03 − 0.02 1 4.BHS 0.08* 0.11** − 0.01 1 5.RRS − 0.03 0.07* − 0.21** 0.32** 1 6.SAI 0.09** 0.13** − 0.15** 0.54** 0.49** 1 7.TAI 0.07* 0.13** − 0.14** 0.57 0.52** 0.84** 1 8.BDI 0.02 0.10** − 0.18** 0.47** 0.67** 0.52** 0.55** 1 9.SA 0.09** 0.14** − 0.14** 0.28** 0.40** 0.32** 0.32** 0.47** 1 10.NSSI − 0.02 0.07* − 0.11** 0.28** 0.39** 0.27** 0.30** 0.43** 0.46** 1 M 16.58 – – 6.83 38.74 43.73 45.24 11.09 – – SD 1.01 – – 3.60 11.99 9.06 8.36 9.54 – – Open in a new tab BHS Beck Hopeless Scale, RRS Ruminative Responses Scale, SAI State Anxiety Inventory, TAI Trait Anxiety Inventory, BDI Beck Depression Inventory, SA Suicidal attempts, NSSI Non-suicidal self-injury, N Number of participants * p < 0.05, ** p < 0.01, *** p < 0.001 Rumination–hopelessness latent profile analysis Latent profiles were analyzed sequentially from 1 classification using the brooding dimension and the reflection dimension of the RRS scale and the feelings about the future dimension and the future expectations dimension of the BHS scale as observational variables, and the indicators of each model fit are shown in Table 2 . And we can see the average posterior probabilities for latent profiles in Table 3 . The results showed that the AIC, BIC, and SSA-BIC values gradually decreased with each additional classification. Starting with three classifications, the entropy values were all greater than 0.8, indicating that the classification accuracy of three, four, and five classifications were all greater than 90%. Considering that profiles need to represent a minimum of 5% of the overall sample, and taking into account model simplicity and interpretability, the 3-classification model was the best-fitting model (Table 2 —model 3, Fig. 1 ). The average posterior probabilities for all three latent classes exceeded 0.90, indicating good classification quality (Table 3 ). Class 1 was characterized by low levels of rumination and average levels of hopelessness, and was labeled the “low rumination–low hopelessness” group ( n = 343, 36.07%). Class 2 exhibited average levels of both rumination and hopelessness, and was labeled the “moderate rumination–moderate hopelessness” group ( n = 531, 55.84%). Class 3 was marked by high levels of both rumination and hopelessness, and was labeled the “high rumination–high hopelessness” group ( n = 77, 8.09%). Table 4 shows the differences in the scores of the three categories on the observed variables. ANOVA results revealed significant differences ( p < 0.05) among the groups on the RRS brooding, RRS reflection, BHS feelings about the future, and BHS future expectations variables. Post hoc analyses showed that the high rumination–high hopelessness group scored higher than the other two groups on each item. The moderate rumination–moderate hopelessness group scored higher than the low rumination-low hopelessness group on all three observed variables (RRS brooding, RRS reflection, and BHS future expectations) but did not differ from the low rumination-low hopelessness group on BHS feelings about the future. Table 2. Rumination–hopelessness latent profile analysis fitting indicators ( n = 951) Model AIC BIC SSA-BIC Entropy LMR ( p ) BLRT ( p ) Proportion of potential categories Group size 1 10,807.28 10,846.14 10,820.74 – – – – – 2 10,197.56 10,260.71 10,219.42 0.76 < 0.001 < 0.001 0.45/0.55 424/527 3 9540.31 9627.74 9570.58 0.90 < 0.001 < 0.001 0.36/0.56/0.08 343/531/77 4 9404.42 9516.14 9443.10 0.90 0.03 < 0.001 0.34/0.10/0.04/0.52 322/99/35/495 5 9296.40 9432.42 9343.49 0.89 0.02 < 0.001 0.23/0.07/0.40/0.27/0.03 223/69/370/257/32 Open in a new tab AIC Akaike Information Criterion, BIC Bayesian Information Criterion, SSA-BIC sample size-adjusted BIC, LRT Lo-Mendell-Rubin likelihood ratio test, BLRT Bootstrap likelihood ratio test, Entropy information entropy Table 3. Average posterior probabilities for latent profiles Model Proportion of potential categories Average posterior probabilities 1 – – 2 0.45/0.55 Class 1 = 0.93; Class 2 = 0.94 3 0.36/0.56/0.08 Class 1 = 0.96; Class 2 = 0.96; Class 3 = 0.94 4 0.34/0.10/0.04/0.52 Class 1 = 0.96; Class 2 = 0.89; Class 3 = 0.95; Class 4 = 0.94 5 0.23/0.07/0.40/0.27/0.03 Class 1 = 0.88; Class 2 = 0.93; Class 3 = 0.93; Class 4 = 0.96; Class 5 = 0.98 Open in a new tab Fig. 1. Open in a new tab Latent profiles of rumination–hopelessness Table 4. Differences in means of latent profiles across observed variables Observed variables Low rumination–low hopelessness Moderate rumination–moderate hopelessness High rumination–high hopelessness F Post hoc η p 2 Brooding − 1.01 (0.02) 0.35 (0.02) 2.05 (0.07) 1637.73*** (3) > (2) > (1) 0.78 Reflection − 1.02 (0.02) 0.37 (0.02) 1.96 (0.08) 1524.80*** (3) > (2) > (1) 0.76 Feelings about the future − 0.08 (0.05) − 0.02 (0.04) 0.48 (0.12) 10.20*** (3) > (2), (1) 0.02 Future expectations − 0.29 (0.05) 0.07 (0.04) 0.78 (0.12) 42.95*** (3) > (2) > (1) 0.08 Open in a new tab *** p < 0.001 Covariate effects on latent profiles R3STEP was used to examine whether covariates were associated with latent profiles. Using Class 2 as reference, gender significantly differentiated Class 1 from Class 2, indicating that males were more likely to be assigned into Class 1 (Table 5 ). Table 5. R3STEP analysis of covariates with class 2 as reference Class 2 as reference Class 1 Class 3 β SE OR (95%CI) β SE OR (95%CI) Age 0.001 0.07 1.00 (0.89, 1.13) − 0.16 0.14 0.86 (0.68, 1.08) Gender 0.94*** 0.16 2.56 (1.98, 3.30) − 0.46 0.30 0.63 (0.39, 1.03) Grade − 0.08 0.16 0.92 (0.70, 1.20) 0.46 0.32 1.59 (0.94, 2.69) BDI − 0.17*** 0.02 0.84 (0.82, 0.87) 0.18*** 0.02 1.19 (1.15, 1.24) TAI − 0.10*** 0.01 0.90 (0.89, 0.92) 0.23*** 0.04 1.26 (1.19, 1.34) SAI − 0.08*** 0.01 0.92 (0.91, 0.94) 0.22*** 0.03 1.25 (1.19, 1.32) Open in a new tab BDI Beck Depression Inventory, TAI Trait Anxiety Inventory, SAI State Anxiety Inventory Class 1 = low rumination–low hopelessness group, Class 2 = moderate rumination–moderate hopelessness group, Class 3 = high rumination–high hopelessness group. * p < 0.05, ** p < 0.01, *** p < 0.001. OR, Odds Ratio (OR > 1: Increased risk; OR < 1: Decreased risk; OR = 1: No effect) For affective disorders, higher levels of depression, trait anxiety, and state anxiety decreased the likelihood of being assigned to Class 1 relative to Class 2, but increased the likelihood of being assigned to Class 3 relative to Class 2 (Table 5 ). To evaluate the robustness of the findings, we repeated the analyses using a four-class model, which also showed good fit (Table 2 ). The R3STEP analysis from the four-class model indicated that gender and all affective symptoms (depression, trait anxiety, and state anxiety) were associated with profile membership in patterns consistent with those observed in the three-class model (Table S5). Results of BCH pairwise comparisons The BCH approach was applied to examine the associations of the rumination–hopelessness profiles with SA and NSSI, accounting for classification uncertainty. Pairwise comparisons of the predicted probabilities across the three classes showed that all contrasts were significant (Table 6 ). The probability of both SA and NSSI increased monotonically across profiles, following the pattern Class 1 < Class 2 < Class 3. Table 6. BCH analysis for probabilities of SA/NSSI SA/NSS I = 1 Probability (95% CI) OR (95% CI) Class 1 vs. Class 2 Class 1 vs. Class 3 Class 2 vs. Class 3 SA Class 1 = 0.04 (0.02, 0.06) 0.37*** (0.21, 0.63) 0.03*** (0.02, 0.06) 0.09*** (0.06, 0.15) Class 2 = 0.11 (0.09, 0.14) Class 3 = 0.57 (0.47, 0.67) NSSI Class 1 = 0.003 (− 0.004, 0.01) 0.04*** (0.003, 0.48) 0.004*** (0.00,0.06) 0.12*** (0.07,0.20) Class 2 = 0.07 (0.05, 0.09) Class 3 = 0.40 (0.30, 0.49) Open in a new tab SA Suicidal attempts, NSSI Non-suicidal self-injury Class 1 = low rumination–low hopelessness group, Class 2 = moderate rumination–moderate hopelessness group, Class 3 = high rumination–high hopelessness group. * p < 0.05, ** p < 0.01, *** p < 0.001. OR, Odds Ratio (OR > 1: Increased risk; OR < 1: Decreased risk; OR = 1: No effect) On the BCH approach was also applied to four-class sensitivity model. Pairwise comparisons of probabilities revealed a clear severity gradient across groups (Table S6). Class 1 showed the lowest probability of SA, significantly lower than Classes 2 and 3, and its NSSI probability was lower than Class 3. Class 2 demonstrated higher SA and NSSI probabilities than Class 4, but lower than Class 3. Class 3 exhibited the highest probabilities of both SA and NSSI compared with all other classes. Hierarchical logistic regression of latent classes on SA and NSSI Model 2 significantly improved model fit compared to Model 1 (Δχ² (2) = 180.20, p < 0.001), indicating that depression and anxiety were significantly associated with SA. Model 3 also showed improvement over Model 2 (Δχ² (2) = 6.23, p = 0.04), suggesting that the rumination–hopelessness latent categories contributed additional explanatory power for SA. Age, grade, and gender were used as covariates in predicting SA. The results of the hierarchical model for SA are shown in Table 7 , with estimates reported as β coefficients, standard errors, and odds ratios with 95% confidence intervals. In Model 1, grade (β = 1.03, SE = 0.29, OR = 2.79, 95%CI 1.59–4.90, p < 0.001) and gender (β = − 0.91, SE = 0.22, OR = 0.40, 95%CI 0.26–0.62, p < 0.001) were significantly associated with SA, and the difference in age did not reach statistical significance ( p = 0.09). Females exhibited a higher prevalence of SA than males, and a higher level of grade was associated with an increased incidence of SA. Model 2 with the addition of state anxiety and depression found that state anxiety (β = 0.07, SE = 0.02, OR = 1.07, 95%CI 1.03–1.12, p < 0.001) and depression (β = 0.11, SE = 0.01, OR = 1.12, 95%CI 1.09–1.15, p < 0.001) were significantly associated with SA. Higher levels of anxiety and depression were associated with an increased risk of SA. Dummy variables were created for the rumination–hopelessness latent profiles (D1: 0 = moderate rumination–moderate hopelessness, 1 = low rumination–low hopelessness; D2: 0 = moderate rumination–moderate hopelessness, 1 = high rumination–high hopelessness). After controlling for variables in the first two steps, the high rumination–high hopelessness group (β = 0.84, SE = 0.34, OR = 2.31, 95%CI 1.20–4.45, p = 0.01, p _FDR = 0.02) showed a significantly higher risk of SA compared to the moderate rumination–moderate hopelessness group. No significant difference was found between the low rumination–low hopelessness group and the moderate rumination–moderate hopelessness group ( p = 0.63, p_ FDR = 0.63). In Model 3, age (β = 0.22, SE = 0.11, OR = 1.25, 95%CI 1.02–1.54, p = 0.03, p _FDR = 0.046) was significant associated with SA, while the effect of gender was no longer statistically significant ( p = 0.05, p _FDR = 0.06). The regression coefficients are presented in Fig. 2 . Table 7. Hierarchical regression analysis of SA and NSSI (Class 2 as reference) Model Variables SA NSSI β (SE) OR (95%CI) P/ P_ FDR β (SE) OR (95%CI) P/ P_ FDR Model 1 Age 0.16 (0.09) 1.17 (0.98, 1.41) 0.09 − 0.18 (0.14) 0.84 (0.64, 1.10) 0.20 Grade 1.03 (0.29) 2.79 (1.59, 4.90) < 0.001 0.77 (0.33) 2.16 (1.13, 4.11) 0.02 Gender − 0.91 (0.22) 0.40 (0.26, 0.62) < 0.001 − 0.93 (0.29) 0.40 (0.23, 0.70) 0.001 Model 2 Age 0.20 (0.10) 1.23 (1.00, 1.50) 0.05 − 0.24 (0.16) 0.79 (0.57, 1.09) 0.15 Grade 0.85 (0.33) 2.34 (1.22, 4.50) 0.01 0.54 (0.39) 1.72 (0.80, 3.68) 0.16 Gender − 0.51 (0.26) 0.60 (0.36, 0.998) 0.049 − 0.52 (0.33) 0.60 (0.31, 1.15) 0.12 Depression 0.11 (0.01) 1.12 (1.09, 1.15) < 0.001 0.13 (0.02) 1.14 (1.10, 1.18) < 0.001 Anxiety 0.07 (0.02) 1.07 (1.03, 1.12) < 0.001 0.05 (0.02) 1.05 (1.00, 1.10) 0.04 Model 3 Age 0.22 (0.11) 1.25 (1.02, 1.54) 0.03/0.046 − 0.20 (0.17) 0.82 (0.59, 1.14) 0.23/ 0.31 Grade 0.85 (0.34) 2.35 (1.21, 4.57) 0.01/0.02 0.55 (0.39) 1.74 (0.81, 3.74) 0.16/ 0.31 Gender − 0.52 (0.27) 0.60 (0.35, 1.01) 0.05/0.06 − 0.32 (0.34) 0.72 (0.37, 1.41) 0.34/ 0.34 Depression 0.11 (0.02) 1.11 (1.08, 1.14) < 0.001/< 0.001 0.12 (0.02) 1.12 (1.08, 1.17) < 0.001/ < 0.001 Anxiety 0.06 (0.02) 1.06 (1.02, 1.11) 0.004/0.01 0.04 (0.03) 1.04 (0.989, 1.09) 0.13/ 0.31 D1 0.16 (0.34) 1.18 (0.60, 2.30) 0.63/0.63 − 1.51 (0.76) 0.22 0.05, 0.98) 0.046/ 0.194 D2 0.84 (0.34) 2.31 (1.20, 4.45) 0.01/0.02 0.43 (0.37) 1.54 (0.74, 3.19) 0.25/ 0.31 Open in a new tab SA Suicidal attempts, NSSI Non-suicidal self-injury Class 2 = moderate rumination–moderate hopelessness group, Categorical variable coding (D1: 0 = moderate rumination–moderate hopelessness group, 1 = low rumination–low hopelessness group; D2: 0 = moderate rumination–moderate hopelessness group, 1 = high rumination–high hopelessness group). OR, Odds Ratio (OR > 1: Increased risk; OR < 1: Decreased risk; OR = 1: No effect) Fig. 2. Open in a new tab Odds Ratios from hierarchical logistic regression analyses. SA Suicidal attempts, NSSI Non-suicidal self-injury. Categorical variable coding (D1: 0 = moderate rumination–moderate hopelessness group, 1 = low rumination–low hopelessness group; D2: 0 = moderate rumination–moderate hopelessness group, 1 = high rumination–high hopelessness group). OR, Odds Ratio (OR > 1: Increased risk; OR < 1: Decreased risk; OR = 1: No effect) Model 2 significantly improved model fit compared to Model 1(Δχ² (2) = 139.40, p < 0.001), indicating that depression and anxiety showed significant associations with NSSI. Model 3 also showed improvement over Model 2 (Δχ² (2) = 7.42, p = 0.02), suggesting that the rumination–hopelessness latent classes contributed additional explanatory power for NSSI. Age, grade, and gender were included as covariates in the NSSI model. The hierarchical model results examining associations with NSSI were presented in Table 7 . In Model 1, both grade (β = 0.77, SE = 0.33, OR = 2.16, 95%CI 1.13–4.11, p = 0.02) and gender (β = − 0.93, SE = 0.29, OR = 0.40, 95%CI 0.23–0.70, p = 0.001) were significantly associated with NSSI. However, after including additional variables in Models 2 and 3, the associations of grade and gender with NSSI were no longer statistically significant. In Model 2, both depression (β = 0.13, SE = 0.02, OR = 1.14, 95%CI 1.10–1.18, p < 0.001) and anxiety (β = 0.05, SE = 0.02, OR = 1.05, 95%CI 1.00–1.10, p = 0.04) were significantly associated with NSSI. In Model 3, depression showed significant association with NSSI, while the effect of anxiety was no longer statistically significant ( p = 0.13, p _FDR = 0.31). Neither the low rumination–low hopelessness group (D1: p = 0.046, p _FDR = 0.19) nor the high rumination–high hopelessness group (D2: p = 0.25, p _FDR = 0.31) showed significant differences with the moderate rumination–moderate hopelessness group on NSSI. The regression coefficients are presented in Fig. 2 . We also conducted sensitivity analyses of TAI. In the hierarchical logistic regression analyses, we replaced SAI with TAI to evaluate the stability of the associations when using an alternative anxiety construct. The results were highly consistent with the main analyses, indicating that the inclusion of trait anxiety did not materially alter the interpretation of group differences (Table S3). Because the four-class also demonstrated good fit and competitive performance, we re-estimated four-class regression model. Sensitivity analysis identified that Class 3 had a significantly higher risk of SA compared with Class 4. In contrast, the differences between Class 4 and the other profiles on NSSI were not statistically significant. This pattern may reflect the reduced between-class contrasts that occur when the number of classes increases, as shown in Table S7. The overall direction of effects remained consistent across models. Discussion Rumination and hopelessness are closely associated with affective disorders and self-harm behaviors. Our study used latent profile analysis (LPA) to examine the generalized association of rumination–hopelessness profiles across different types of self-harm behaviors in a non-clinical adolescent sample. Our results reveal distinct latent profiles characterized by combined features of rumination and hopelessness exhibited differential effects related to SA and NSSI, emphasizing the importance of utilizing LPA to capture complex cognitive co-occurrence. More specifically, the high rumination–high hopelessness profile was uniquely associated with a highly likelihood of SA. This association remained significant after statistically controlling for general affective symptoms. In contrast, NSSI was significantly linked to the severity of depressive symptoms, indicating less specificity in its association with the cognitive latent profiles. Latent profiles of rumination–hopelessness among adolescents and their characteristics Our study identified three latent profiles through LPA: the low rumination–low hopelessness group, the moderate rumination–moderate hopelessness group, and the high rumination–high hopelessness group. These groups exhibited significant and distinct step-wise differences across the severity of rumination and hopelessness, confirming that the combination of these cognitive traits varies meaningfully in our adolescent sample. The clustering of high levels of rumination and hopelessness in the same profile is consistent with prior research showing their frequent co-occurrence in adolescents. This interrelationship may be explained by the reciprocal relationship between rumination and hopelessness. Individuals tended to engage in persistent and repetitive thinking about hopelessness-related cognitive content. This sustained focus on hopelessness may trigger further rumination, creating a feedback loop that exacerbates both constructs [ 33 ]. Additionally, rumination and hopelessness may share similar neural mechanisms. Rumination has been linked to increased functional connectivity within emotional networks, particularly positive connectivity between the amygdala and prefrontal regions. This connectivity may influence emotional regulation through comparable neural circuits [ 34 ]. In another study, depressed patients with a history of suicide were found to have significant correlations between activation in the orbital frontal cortex and the dorsolateral prefrontal cortex. This activation was associated with feelings of hopelessness and may be similar to rumination, involving the over processing of negative emotional information in the prefrontal cortex [ 35 ]. Interpretation of covariate and distal outcome analyses Our analysis using the R3STEP approach revealed important findings about the factors influencing membership in the identified latent profiles. We observed a significant gradient in profile assignment based on affective symptoms: individuals with lower levels of anxiety and depression were significantly more likely to belong to the low rumination–low hopelessness profile, while those with higher affective distress were more likely to be assigned to the high rumination–high hopelessness profile (Class 3). This pattern suggests that affective distress acts as a trigger for the formation of the high-risk cognitive profile, fostering the simultaneous enhancement of rumination and hopelessness [ 36 – 38 ]. Furthermore, we found that gender significantly influenced profile assignment, with males being more likely to be classified into the lower-risk group (Class 1) relative to the moderate-risk group (Class 2). This is consistent with existing literature indicating that adolescent females generally exhibit higher levels of rumination [ 39 ], supporting that established demographic differences in key cognitive traits contribute to the divergence into distinct self-harm risk profiles. Using the BCH method to account for classification uncertainty, we established a clear link between the latent cognitive profiles and the distal outcomes of SA and NSSI. The probability of engaging in both SA and NSSI increased progressively and significantly across the profiles, following a precise severity gradient (Class 1 < Class 2 < Class 3). This finding supports previous research demonstrating that high levels of rumination and hopelessness are individually associated with heightened self-harm risk [ 15 , 16 , 18 , 19 ]. Crucially, our study extends this by demonstrating that the co-occurrence of high rumination and high hopelessness forms a combined, high-risk cognitive profile that is systematically linked to the heightened likelihood of self-harm behaviors. This systematic gradient underscores the clinical utility of identifying this specific, extreme cognitive signature when assessing self-harm risk in adolescents. Specific effects of rumination–hopelessness latent profiles on SA and NSSI Our hierarchical logistic regression analyses revealed that the latent profiles of rumination and hopelessness had different effects on SA and NSSI. The most compelling finding is the robust association between the high rumination–high hopelessness profile and SA risk, independent of affective distress. Previous meta-analytic evidence indicated that overall rumination and brooding rumination were moderately positively associated with SA, whereas the reflection showed no significant association with SA [ 40 ]. In our study, differences in both brooding and reflection dimensions were observed across latent profiles, suggesting that overall rumination may be associated with SA. Both retrospective and prospective analyses have consistently found that levels of hopelessness were significantly higher in individuals with a history of suicide behaviors compared to those without such a history [ 41 ]. Our study extends these findings by characterizing the combination of high rumination and high hopelessness as a dual cognitive constriction in SA. Individuals with co-occurrence of dual cognitive rigidity and despair may develop a tunnel vision, perceiving death as the only viable solution to their crisis [ 42 ], ultimately leading to the escalation of more lethal forms of self-harm such as SA [ 43 ]. The specific cognitive vulnerability in SA extends beyond general affective symptoms, implying that it may serve as a mechanism that translates affective distress into suicidal behaviors [ 44 ]. In contrast to SA, the likelihood of NSSI in our sample was strongly associated with the severity of depressive symptoms, but not uniquely related to the complex cognitive profiles. This finding supports the dominant theoretical understanding of NSSI. NSSI was considered a “hidden” form of self-harm behaviors that occurred below the surface of the iceberg—common in the community but often unreported and without help-seeking behaviors—and typically served as a means for adolescents to quickly escape pain or regulate emotions [ 43 ]. Because the primary goal of NSSI is immediate relief, its occurrence may be more closely tied to the intensity of general negative affect captured by the continuous depressive symptom scores rather than the temporally rigid, future-oriented hopelessness that characterizes high-risk SA profiles. This highlights a fundamental functional and motivational divergence between the two forms of self-harm [ 45 ]. Moreover, the significant effect of grade observed in our analyses suggested that the risk of suicide behaviors was higher among students in upper grades, potentially reflecting both ongoing cognitive development and increasing academic pressure associated with higher educational levels. This is consistent with previous evidence indicating a strong correlation between academic stress and self-harm behaviors, mediated in part by internalizing symptoms such as anxiety and depression [ 46 ]. Limitations This study has several limitations that warrant cautious interpretation of the findings. Firstly, the cross-sectional nature of the data restricts our ability to establish temporal or causal relationships. Future studies using clinician-administered assessments and longitudinal designs are needed to further validate the predictive utility of these profiles across different developmental and clinical contexts. Secondly, although we demonstrated the suitability of the Chinese-adapted measures, the reliance on self-report and single-item assessment for SA and NSSI may introduce misclassification bias. Future studies are encouraged to consider the use of multi-item scales to improve measurement fidelity and diagnostic accuracy. Thirdly, while formal testing suggested common method bias was not severe, we recognize that method variance remains a potential limitation, possibly leading to inflated associations between self-report constructs. Future studies should employ multi-modal assessment strategies to ensure the independence and accurate estimation. Moreover, the use of a non-clinical adolescent sample limits the generalizability of the findings to clinical populations, where self-harm rates and comorbidity levels are substantially higher. Conclusion Our study employed a person-centered LPA approach to explore combined features of rumination and hopelessness, two cognitive factors closely associated with adolescent self-continuity. The findings suggest that a specific, severe cognitive profile (high rumination–high hopelessness) is robustly and uniquely associated with SA, whereas NSSI is primarily related to the intensity of depressive symptoms. Our study resolves a key theoretical gap in self-harm mechanism by mapping cognitive phenotypes to differential self-harm pathways. Moreover, our work indicates actionable taxonomy that enables pre-symptomatic screening via rumination/hopelessness assessment and reorients intervention toward cognitive-specific targets alongside traditional affective management. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1. (1.7MB, docx) Author contributions H.A. and Y.L.: Conceptualization, funding acquisition. M.W. and H.A.: formal analysis, writing—original draft, visualization. S.C. and J.G.: investigation, data curation. C.C. and H.A.: writing—review & editing, validation. All authors: result interpretation, final approval of the manuscript. Funding This study is funded by the National Natural Science Foundation of China (32371104), theMajor Project of the National Social Science Foundation (20&ZD153), Shenzhen-Hong KongInstitute of Brain Science-Shenzhen Fundamental Research Institutions (2023SHIBS0004), Project (PT-2024-06B) from the National Human Genetic Resources Sharing Service Platform(2005DKA21300), and Tianjin Mental Health Education Research Special Project-CommissionedResearch Topic (2025WTGX03). Data availability Due to ethical restrictions, the data are not publicly available. Anonymized data may be shared upon reasonable request and institutional approval. Declarations Ethics approval and consent to participate This study received ethical approval from the university's ethical review board (TJUE-2024-544). Prior to conducting the study, informed consent was obtained from all the adolescents and their primary caregivers. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References 1. Nock MK. Self-injury. Ann Rev Clin Psychol. 2010;6(1):339–63. [ DOI ] [ PubMed ] [ Google Scholar ] 2. WHO, Suicide. 2025. https://www.who.int/news-room/fact-sheets/detail/suicide . Accessed on 2 Dec 2025. 3. Kim S, Park J, Lee H, Lee H, Woo S, Kwon R, Kim S, Koyanagi A, Smith L, Rahmati M, et al. Global public concern of childhood and adolescence suicide: a new perspective and new strategies for suicide prevention in the post-pandemic era. World J Pediatr. 2024;20(9):872–900. [ DOI ] [ PubMed ] [ Google Scholar ] 4. Xiao Q, Song X, Huang L, Hou D, Huang X. Global prevalence and characteristics of non-suicidal self-injury between 2010 and 2021 among a non-clinical sample of adolescents: A meta-analysis. Front Psychiatry. 2022;13:912441. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 5. Qu D, Wen X, Liu B, Zhang X, He Y, Chen D, Duan X, Yu J, Liu D, Zhang X, et al. Non-suicidal self-injury in Chinese population: a scoping review of prevalence, method, risk factors and preventive interventions. Lancet Reg Health West Pac. 2023;37:100794. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 6. Yu G, Huang X. A comparison of the rates of prevalence of students’ mental health problems: evidence based on Meta-analysis. Educ Res. 2023;44:105–21. [ Google Scholar ] 7. Branje S, de Moor EL, Spitzer J, Becht AI. Dynamics of identity development in adolescence: A decade in review. J Res Adolesc. 2021;31(4):908–27. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 8. Barredo J, Bozzay ML, Primack JM, Schatten HT, Armey MF, Carpenter LL, Philip NS. Translating interventional neuroscience to suicide: it’s about time. Biol Psychiatry. 2021;89(11):1073–83. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 9. Freichel R, Nock MK, O’Shea BA. A network outcome analysis of psychological risk factors driving suicide risk in emergency department patients. Nat Ment Health. 2025;3(3):1–8. [ Google Scholar ] 10. Hsu KJ, Beard C, Rifkin L, Dillon DG, Pizzagalli DA, Bjorgvinsson T. Transdiagnostic mechanisms in depression and anxiety: the role of rumination and attentional control. J Affect Disord. 2015;188:22–7. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 11. Klonsky ED, May AM. The three-step theory (3ST): A new theory of suicide rooted in the ideation-to-action framework. Int J Cogn Therapy. 2015;8(2):114–29. [ Google Scholar ] 12. Taylor PJ, Jomar K, Dhingra K, Forrester R, Shahmalak U, Dickson JM. A meta-analysis of the prevalence of different functions of non-suicidal self-injury. J Affect Disord. 2018;227:759–69. [ DOI ] [ PubMed ] [ Google Scholar ] 13. Auerbach RP, Pagliaccio D, Allison GO, Alqueza KL, Alonso MF. Neural correlates associated with suicide and nonsuicidal self-injury in youth. Biol Psychiatry. 2021;89(2):119–33. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 14. Marchetti I. Hopelessness: a network analysis. Cogn Therapy Res. 2019;43(3):611–9. [ Google Scholar ] 15. Franklin JC, Ribeiro JD, Fox KR, Bentley KH, Kleiman EM, Huang X, Musacchio KM, Jaroszewski AC, Chang BP, Nock MK. Risk factors for suicidal thoughts and behaviors: A meta-analysis of 50 years of research. Psychol Bull. 2017;143(2):187–232. [ DOI ] [ PubMed ] [ Google Scholar ] 16. Perez Rodriguez S, Marco Salvador JH, Garcia-Alandete J. The role of hopelessness and meaning in life in a clinical sample with non-suicidal self-injury and suicide attempts. Psicothema. 2017;29(3):323–8. [ DOI ] [ PubMed ] [ Google Scholar ] 17. Nolen-Hoeksema S, Wisco BE, Lyubomirsky S. Rethinking rumination. Perspect Psychol Sci. 2008;3(5):400–24. [ DOI ] [ PubMed ] [ Google Scholar ] 18. Le GH, Wong S, Au H, Badulescu S, Gill H, Vasudeva S, Teopiz KM, Rhee TG, Ho R, Kwan ATH, et al. Association between rumination, suicidal ideation and suicide attempts in persons with depressive and other mood disorders and healthy controls: A systematic review and meta-analysis. J Affect Disord. 2025;368:513–27. [ DOI ] [ PubMed ] [ Google Scholar ] 19. He K, Chen L, Fan H, Sun L, Yang T, Liu H, Wang K. Effects of rumination and emotional regulation on Non-Suicidal Self-Injury behaviors in depressed adolescents in china: A multicenter study. Psychol Res Behav Manag. 2025;18:271–9. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 20. Hilt LM, Cha CB, Nolen-Hoeksema S. Nonsuicidal self-injury in young adolescent girls: moderators of the distress-function relationship. J Consult Clin Psychol. 2008;76(1):63–71. [ DOI ] [ PubMed ] [ Google Scholar ] 21. Gong T, Ren Y, Wu J, Jiang Y, Hu W, You J. The associations among self-criticism, hopelessness, rumination, and NSSI in adolescents: A moderated mediation model. J Adolesc. 2019;72:1–9. [ DOI ] [ PubMed ] [ Google Scholar ] 22. Tang P, Kostyrka-Allchorne K, Butura AM, Phillips-Owen J, Sonuga-Barke E. Reciprocal developmental pathways between future-related thinking and symptoms of adolescent depression and anxiety: A systematic review and meta-analysis of longitudinal studies. Clin Psychol Rev. 2024;112:102465. [ DOI ] [ PubMed ] [ Google Scholar ] 23. Zhang Y, Xu W, McDonnell D, Wang JL. The relationship between childhood maltreatment subtypes and adolescent internalizing problems: the mediating role of maladaptive cognitive emotion regulation strategies. Child Abuse Negl. 2024;152:106796. [ DOI ] [ PubMed ] [ Google Scholar ] 24. Yang J, Ling Y, Xiao J, Yao S-Q. The Chinese version of ruminative responses scale in high school students: its reliability and validity. Chin J Clin Psychol. 2009;17:27–31. [ Google Scholar ] 25. Han X, Yang H. Chinese version of Nolen-Hoeksema ruminative responses scale (RRS) used in 912 college students: reliability and validity. Chin J Clin Psychol. 2009;17(5):550–549. [ Google Scholar ] 26. Treynor W, Gonzalez R, Nolen-Hoeksema S. Rumination reconsidered: a psychometric analysis. Cogn Therapy Res. 2003;27(3):247–59. [ Google Scholar ] 27. Kong Y, Zhang J, Jia S, Zhou L. Reliability and validity of the Beck hopelessness scale for adolescent. Chin Mental Health J. 2007;21(10):686–9. [ Google Scholar ] 28. Shek DT. The Chinese version of the State-Trait anxiety inventory: its relationship to different measures of psychological well-being. J Clin Psychol. 1993;49(3):349–58. [ DOI ] [ PubMed ] [ Google Scholar ] 29. Song X, Lu C, Yu Y, Wan Y, Ma F, Chen L, Wang Y. Investigation of the reliability and validity of the Chinese version of the Beck depression Inventory-Ⅱ in middle school students of Yi ethnicity in Liangshan Prefecture, China. Sichuan Da Xue Xue Bao Yi Xue Ban. 2024;55(6):1567–72. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 30. Nock MK, Green JG, Hwang I, McLaughlin KA, Sampson NA, Zaslavsky AM, Kessler RC. Prevalence, correlates, and treatment of lifetime suicidal behavior among adolescents: results from the National comorbidity survey replication adolescent supplement. JAMA Psychiatry. 2013;70(3):300–10. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 31. Wen Z, Xie J, Wang H. Principles, procedures and programs of latent class models. J East China Normal Univ (Educational Sciences). 2023;41(1):1. [ Google Scholar ] 32. Asparouhov T, Muthén B. Auxiliary variables in mixture modeling: using the BCH method in Mplus to estimate a distal outcome model and an arbitrary secondary model. Los Angeles, CA; 2021. 33. Starr LR, Davila J. Responding to anxiety with rumination and hopelessness: mechanism of anxiety-Depression symptom Co-Occurrence? Cognit Ther Res. 2012;36(4):321–37. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 34. Feurer C, Jimmy J, Chang F, Langenecker SA, Phan KL, Ajilore O, Klumpp H. Resting state functional connectivity correlates of rumination and worry in internalizing psychopathologies. Depress Anxiety. 2021;38(5):488–97. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 35. Kamimura H, Matsuoka T, Okai H, Shimizu N, Harada S, Matsuo K. The associations between suicide-related behaviors, prefrontal dysfunction in emotional cognition, and personality traits in mood disorders. Sci Rep. 2022;12(1):17377. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 36. Pedersen H, Gronnaess I, Bendixen M, Hagen R, Kennair LEO. Metacognitions and brooding predict depressive symptoms in a community adolescent sample. BMC Psychiatry. 2022;22(1):157. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 37. Rood L, Roelofs J, Bogels SM, Alloy LB. Dimensions of negative thinking and the relations with symptoms of depression and anxiety in children and adolescents. Cognit Ther Res. 2010;34(4):333–42. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 38. Cunningham S, Gunn T, Alladin A, Cawthorpe D. Anxiety, depression and hopelessness in adolescents: a structural equation model. J Can Acad Child Adolesc Psychiatry. 2008;17(3):137–44. [ PMC free article ] [ PubMed ] [ Google Scholar ] 39. Dawson GC, Adrian M, Chu P, McCauley E, Vander Stoep A. Associations between Sex, Rumination, and depressive symptoms in late adolescence: A Four-Year longitudinal investigation. J Clin Child Adolesc Psychol. 2023;52(5):675–85. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 40. Rogers ML, Joiner TE. Rumination, suicidal ideation, and suicide attempts: A meta-analytic review. Rev Gen Psychol. 2017;21(2):132–42. [ Google Scholar ] 41. Qiu T, Klonsky ED, Klein DN. Hopelessness predicts suicide ideation but not attempts: A 10-year longitudinal study. Suicide Life‐Threatening Behav. 2017;47(6):718–22. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 42. Innamorati M, Erbuto D, Rogante E, Sarubbi S, Trocchia MA, Cifrodelli M, Migliorati M, Comparelli A, Berardelli I, Pompili M. Mental States and temperaments contributing to suicidal crisis in psychiatric inpatients: a cross-sectional and validation study. BMC Psychiatry. 2025;25(1):140. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 43. Geulayov G, Casey D, McDonald KC, Foster P, Pritchard K, Wells C, Clements C, Kapur N, Ness J, Waters K, et al. Incidence of suicide, hospital-presenting non-fatal self-harm, and community-occurring non-fatal self-harm in adolescents in England (the iceberg model of self-harm): a retrospective study. Lancet Psychiatry. 2018;5(2):167–74. [ DOI ] [ PubMed ] [ Google Scholar ] 44. Mann JJ, Rizk MM. A Brain-Centric model of suicidal behavior. Am J Psychiatry. 2020;177(10):902–16. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 45. Wolff JC, Thompson E, Thomas SA, Nesi J, Bettis AH, Ransford B, Scopelliti K, Frazier EA, Liu RT. Emotion dysregulation and non-suicidal self-injury: A systematic review and meta-analysis. Eur Psychiatry. 2019;59:25–36. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] 46. Chen H, Guo H, Chen H, Cao X, Liu J, Chen X, Tian Y, Tang H, Wang X, Zhou J. Influence of academic stress and school bullying on self-harm behaviors among Chinese middle school students: the mediation effect of depression and anxiety. Front Public Health. 2022;10:1049051. [ DOI ] [ PMC free article ] [ PubMed ] [ Google Scholar ] Associated Data This section collects any data citations, data availability statements, or supplementary materials included in this article. Supplementary Materials Supplementary Material 1. (1.7MB, docx) Data Availability Statement Due to ethical restrictions, the data are not publicly available. Anonymized data may be shared upon reasonable request and institutional approval. Articles from Child and Adolescent Psychiatry and Mental Health are provided here courtesy of BMC ACTIONS View on publisher site PDF (1.3 MB) Cite Collections Permalink RESOURCES Similar articles Cited by other articles Links to NCBI Databases Follow NCBI NCBI on X (formerly known as Twitter) NCBI on Facebook NCBI on LinkedIn NCBI on GitHub NCBI RSS feed Connect with NLM NLM on X (formerly known as Twitter) NLM on Facebook NLM on YouTube National Library of Medicine 8600 Rockville Pike Bethesda, MD 20894 Web Policies FOIA HHS Vulnerability Disclosure Help Accessibility Careers NLM NIH HHS USA.gov Back to Top

Related documents

• Help-seeking patterns for non-suicidal self-injury across borderline symptom profiles in a large sample of university students.#1107376
• Assessing Repetitive Negative Thinking Across the Adult Lifespan: Validation and Measurement Invariance of the Dutch Persistent and Intrusive Negative Thoughts Scale (PINTS)#439836
• Assessing Repetitive Negative Thinking Across the Adult Lifespan: Validation and Measurement Invariance of the Dutch Persistent and Intrusive Negative Thoughts Scale (PINTS)#16097
• When trauma triggers craving: meta-analytic evidence for the mediating role of negative affect among trauma-exposed individuals who use substances#127508
• Predictors of suicide attempt in adolescents with a history of self-harm: a machine learning study combining psychosocial and actigraphy measures#928488