Front. Public Health Frontiers in Public Health Front. Public Health 2296-2565 Frontiers Media S.A. 10.3389/fpubh.2026.1730857 Original Research The relationship between self-compassion and self-worth in patients with traumatic brain injury: a latent profile analysis and examination of influencing factors Zhang Huijuan † Writing – review & editing Data curation Validation Investigation Software Writing – original draft Zhao Lili † Writing – review & editing Investigation Writing – original draft Conceptualization Project administration Li Ling * Methodology Project administration Investigation Writing – review & editing Conceptualization Funding acquisition Writing – original draft Zhangjiakou First Hospital, Zhangjiakou, Hebei, China Correspondence: Ling Li, l109989765@163.com

These authors have contributed equally to this work and share first authorship

 12 02 2026 2026 14 1730857 23 10 2025 25 12 2025 15 01 2026 Copyright © 2026 Zhang, Zhao and Li. 2026 Zhang, Zhao and Li https://creativecommons.org/licenses/by/4.0/ This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. Background

Traumatic brain injury (TBI) often leads to post-traumatic syndrome, including symptoms such as dizziness, headaches, and tinnitus, significantly compromising patients’ quality of life. Self-compassion and self-worth, as key self-perception factors, have been widely demonstrated in previous research to buffer against negative self-evaluations, enhance self-worth, and improve quality of life. However, the internal heterogeneity of self-compassion and self-worth within the TBI patient population has not been sufficiently explored. This study aimed to identify potential subtypes based on the relationship between self-compassion and self-worth in TBI patients using latent profile analysis and to explore their influencing factors, thereby revealing individual differences and providing a basis for targeted interventions.

 Method

This cross-sectional study employed a convenience sampling method to recruit 586 diagnosed TBI patients between January and May 2025. Participants completed scales assessing self-compassion, self-worth, social support, and death anxiety to evaluate core variables. Latent profile analysis was conducted based on scores from the Self-compassion Scale and the Self-worth Scale to identify distinct profiles. Multinomial logistic regression analysis was used to examine the effects of demographic variables, injury-related factors, and psychological factors on profile membership.

 Result

This study found a significant positive correlation between self-compassion and self-worth in TBI patients. Latent profile analysis identified three distinct profiles: a ‘High self-compassion-high self-worth’ group, a ‘Moderate self-compassion-moderate self-worth’ group, and a ‘Moderate self-compassion-low self-worth’ group. Multinomial logistic regression analysis revealed that social support, death anxiety, and gender were significant factors influencing profile membership.

 Conclusion

This study uncovered heterogeneous profiles characterizing the relationship between self-compassion and self-worth in patients with traumatic brain injury and identified gender, death anxiety, and social support as key influencing factors. These findings underscore the necessity for targeted psychological interventions, such as enhancing social support and reducing death anxiety, to improve self-compassion in TBI patients, thereby fostering self-worth and quality of life.

 death anxiety self-compassion self-worth social support traumatic brain injury patients The author(s) declared that financial support was not received for this work and/or its publication. section-at-acceptance Public Mental Health Introduction

Traumatic brain injury (TBI) refers to damage to the skull, meninges, intracranial vessels, and brain tissue resulting from external mechanical forces (1, 2). Such injuries often induce acute neurological dysfunction, with long-lasting cognitive, emotional, and physical impairments (3, 4). Globally, TBI affects a substantial number of individuals, with disproportionately higher prevalence in low- and middle-income countries, mainly due to traffic accidents, falls, and violence (5, 6). Consequently, research has traditionally emphasized physiological rehabilitation and cognitive training (7, 8). For example, Martínez-Molina, Siponkoski (9) demonstrated that music interventions significantly improved executive functioning among individuals with TBI. In recent years, however, increasing attention has been given to addressing negative psychological factors to further enhance quality of life (8, 10), including symptoms such as anxiety, depression, and fear (11, 12). Nevertheless, insufficient attention has been given to the role of positive psychological resources in recovery—particularly the interaction between self-compassion and self-worth.

Self-compassion refers to a mindful, kind, and non-judgmental stance toward one’s own suffering. It encompasses three central elements: self-kindness, common humanity, and mindfulness (13, 14). In the context of TBI, self-compassion may act as a protective mechanism, supporting patients in coping with identity crises, functional impairment, and social isolation (15, 16). Individuals with higher self-compassion are more likely to approach personal limitations with tolerance, thus reducing anxiety and depression (17). However, due to executive dysfunction and challenges in emotion regulation, many individuals with TBI struggle to develop or sustain self-compassion (18). For example, prefrontal cortex injury may diminish the capacity for self-reflection, amplifying self-criticism and emotional numbness. Recent findings further indicate no significant differences in self-compassion or self-efficacy levels between patients with TBI and healthy peers (19, 20). Yet, the relationship between self-compassion and self-worth in individuals with TBI remains largely unexplored.

Self-worth, derived from conditional self-worth theory, reflects individuals’ sense of personal value based on achievements, social recognition, or intrinsic attributes (21, 22). Serving as a central pillar of mental health, self-worth is particularly sensitive during TBI rehabilitation (23). Traumatic events often destabilize patients’ physical functioning, social roles, and occupational identities, leading to sharp fluctuations in perceived self-worth (24). According to psychosocial development theory (25), self-worth is formed through an interaction between competence perception and social reinforcement. However, patients with TBI frequently suffer diminished self-worth due to functional deficits and societal stigma (26). Research has shown that low self-worth is significantly associated with depression, loneliness, and reduced treatment adherence among TBI survivors (27, 28). Strikingly, little is known about whether self-compassion functions as a preceding factor that bolsters self-worth. Ignoring this link is unfortunate, since higher self-compassion may help patients preserve self-worth through acceptance and self-acknowledgment, while lower self-compassion could exacerbate self-denial and self-criticism (29, 30).

Existing research often employs variable-centered analyses that presume psychological homogeneity within samples. In broader populations and chronic illness contexts, many studies consistently report a positive association between self-compassion and self-worth (31, 32). For instance, self-compassion has been found to enhance self-worth in women with chronic pain (33). Yet, some studies suggest weaker or domain-specific links, such as between self-compassion and appearance-contingent self-worth (34). In depressed women, self-compassion indirectly enhances self-worth through mechanisms such as self-acceptance and psychological capital (35). These inconsistencies may stem from not accounting for subgroup heterogeneity. For individuals with TBI, physical impairments and role disruption often interact with negative affect and social feedback, threatening both self-worth and self-compassion (26, 36). Conversely, heightened self-compassion promotes gentle self-care, perspective-taking on shared human suffering, and mindful nonjudgment—which collectively suppress self-denigration and catastrophizing, leading to stabilized self-worth, stronger adherence to rehabilitation, and improved social functioning (37, 38). Given the clinical diversity of TBI in injury severity, treatment burden, financial strain, and stigma, person-centered analytic approaches such as latent profile analysis (LPA) may uncover more nuanced profiles of self-compassion and self-worth compared with variable-centered models.

From the stress-buffering model of social support, social networks provide emotional, informational, and instrumental support, shaping both self-perception and emotion regulation (39, 40). For patients with TBI, however, trauma often disrupts social networks and strips away role functions (41), amplifying isolation and crises of self-worth. High levels of social support validate individuals’ limitations and reduce self-criticism, fostering greater self-kindness (42, 43). At the same time, social support helps integrate traumatic experiences into life narratives, rebuild positive identity, and reinforce self-worth (44). Thus, social support is likely to be a key contextual factor influencing profiles of self-compassion and self-worth.

A further crucial factor is death anxiety. Rooted in terror management theory, death anxiety emerges from awareness of mortality, which individuals buffer through bolstering cultural worldviews and self-worth (45). For patients with TBI—who often face sudden bodily vulnerability, functional limitations, and uncertainty regarding recovery—death-related concerns may become salient and may undermine adaptive self-related processes (46, 47). Specifically, heightened death-related thoughts have been shown to predict greater depressive symptoms among individuals with low self-worth, suggesting that mortality concerns can erode self-evaluations when self-worth resources are insufficient (48, 49). Conceptually, elevated death anxiety may also promote threat-focused (50), self-critical coping styles that are inconsistent with self-compassion, which relies on mindful and tolerant acceptance of human limitations (51, 52). Simultaneously, disruption of self-identity and social roles is frequently reported after TBI and is a recognized challenge to psychosocial functioning, potentially contributing to declines in self-worth (53, 54).

To address these research gaps, the present study applies LPA to identify overlooked psychosocial subgroups among patients with TBI. LPA is a person-centered, model-based clustering technique that identifies latent subgroups within a population based on individuals’ response patterns on continuous indicators (55, 56). Unlike variable-centered approaches that estimate average associations assuming population homogeneity (57, 58), LPA characterizes heterogeneity by estimating profile-specific means and assigning individuals to the most likely profile with quantifiable classification accuracy (e.g., entropy). LPA has been widely employed in clinical research, such as classifying pancreatic cancer patients based on psychological capital and death anxiety (59). Though previously used in TBI to explore heterogeneity in cognition, emotional regulation, and rehabilitation outcomes, LPA has not yet been applied to reveal combined distributions of self-compassion and self-worth.

The primary objective of this study is to identify distinct latent profiles characterizing the relationship between self-compassion and self-worth among TBI patients and to examine the demographic, clinical, and psychosocial factors associated with profile membership. A secondary objective is to provide empirically grounded recommendations for profile-specific psychological interventions.

Based on the theoretical frameworks outlined above, we propose the following hypotheses:

Hypothesis 1: Multiple distinct latent profiles of self-compassion and self-worth exist among TBI patients, reflecting heterogeneous patterns of self-perception. Specifically, we anticipate identifying at least three profiles representing varying levels of self-compassion and self-worth concordance.

Hypothesis 2: Social support serves as a protective factor for profile membership. Patients with higher perceived social support are more likely to belong to profiles characterized by elevated self-compassion and self-worth, whereas those with lower social support are more likely to belong to profiles marked by diminished self-worth.

Hypothesis 3: Death anxiety functions as a risk factor for profile membership. Patients with higher death anxiety are less likely to belong to profiles characterized by high self-compassion and self-worth, given that mortality salience may undermine self-compassionate coping and destabilize self-worth.

This study addresses three critical research questions aligned with these hypotheses:

Do distinct profiles of self-compassion and self-worth exist among TBI patients?

How are these profiles associated with demographic, clinical, and psychosocial characteristics?

How can such profiles inform individualized intervention strategies?

Through the application of LPA, this study aims to refine the psychological classification of TBI populations and provide empirical evidence to support tailored, profile-specific rehabilitation approaches. By linking profile characteristics to specific intervention modalities, we seek to bridge the gap between population-level psychological research and individualized clinical practice.

 Methods Participants Ethical considerations

This study strictly adhered to the principles of the Declaration of Helsinki and its subsequent amendments, and it was approved by the Institutional Review Board of the institution (Approval No.: 2025-LW-31). All participants were provided with detailed information prior to recruitment, including the study objectives, procedures, potential risks and benefits, the voluntary nature of participation, and the right to withdraw at any time. Written informed consent was obtained before inclusion. All data were anonymized and accessible only to the research team.

 Study design

A cross-sectional design was employed to examine the heterogeneity of the relationship between self-compassion and self-worth among TBI patients, as well as associated influencing factors. A cross-sectional design enabled data collection at a single time point to explore associations, rather than causal relationships. Participants were recruited using convenience sampling. Core variables assessed included self-compassion, self-worth, social support, and death anxiety.

In the present study, we focused on death anxiety as a theoretically and clinically salient form of existential threat that is directly implicated in self-worth regulation (terror management theory). While other anxiety domains (e.g., generalized anxiety, panic, post-traumatic stress–related anxiety) are also relevant after TBI, our aim was to examine whether an existentially grounded anxiety construct helps explain heterogeneity in the joint distribution of self-compassion and self-worth. Moreover, to minimize participant burden and avoid symptom exacerbation, we excluded individuals with current severe anxiety or depressive disorders; thus, the study was not designed to model clinical anxiety syndromes but rather to test the role of death anxiety as a profile-related correlate.

 Recruitment procedure

Data were collected between January and May 2025 in the neurosurgery outpatient and inpatient departments of four tertiary Class-A hospitals in Hebei Province, China. First, potential participants were screened through electronic medical records according to inclusion criteria. Next, the study was introduced both orally and via posters. Interested patients underwent eligibility interviews. Those deemed eligible provided informed consent and proceeded to complete the questionnaires, which were available in both paper-based and electronic formats (via QR code) to accommodate participant preference.

 Inclusion and exclusion criteria

Inclusion criteria: Age ≥18 years; Clinically confirmed TBI by neuroimaging with a disease course ≥3 months (to ensure stabilization of acute symptoms); Clear consciousness, adequate comprehension, and the ability to complete questionnaires; No severe communication barriers and capable of providing informed consent. Consciousness and comprehension were assessed using the Glasgow Coma Scale (GCS) (60), which evaluates eye response, verbal response, and motor response, with total scores ranging from 3 to 15. Higher scores indicate better consciousness levels. Patients with a GCS score ≤12 (moderate-to-severe coma) were excluded.

Exclusion criteria: Presence of other neurological disorders (e.g., Alzheimer’s disease, Parkinson’s disease) to minimize confounding; Current severe depressive or anxiety disorders, to prevent symptom exacerbation; Severe cognitive impairment or dysphasia; Refusal to participate or inability to complete the questionnaires. Cognitive function was assessed using the Mini-Mental State Examination (MMSE); patients scoring <24 were excluded (61).

Consciousness and comprehension were screened to ensure participants could provide informed consent and complete self-report questionnaires reliably. Although the Glasgow Coma Scale (GCS) is primarily used in acute care, we used the initial GCS score recorded at hospital admission (retrieved from electronic medical records) as a proxy for initial injury severity and level of consciousness, and excluded patients with GCS ≤ 12 (i.e., a history of moderate-to-severe coma), as such individuals are more likely to present with persistent disorders of consciousness and/or severe cognitive-communication limitations that may compromise questionnaire validity. To additionally verify current cognitive capacity at the time of survey administration (≥3 months post-injury), we applied the Mini-Mental State Examination (MMSE), excluding patients scoring < 24.

 Minimum sample size

Sample size estimation was based on statistical requirements for LPA and multivariate logistic regression. For LPA, referencing Nylund, Asparouhov (62), at least 300–500 participants are recommended to ensure model stability and clarity when assuming 3–5 potential profiles. For logistic regression, using G*power software with parameters set at f2 = 0.15, α = 0.05, 13 predictors, and 95% statistical power, the minimum required sample size was 189. Considering a 20% attrition rate, the minimum viable sample size was calculated as 227. Ultimately, this study recruited 586 participants—well exceeding the threshold and ensuring robust statistical power.

 Study sample

A total of 605 participants were recruited. Of these, 6 declined to provide informed consent, 3 were excluded due to other neurological disorders, and 10 were excluded because of incomplete questionnaires or response bias. The final effective sample comprised 586 participants, yielding an effective rate of 96.85%. Among them, 361 (61.60%) were men and 225 (38.40%) were women. Educational background was primarily middle to high school (N = 245, 41.80%), marital status was predominantly married (N = 336, 57.30%), and the majority resided in urban areas (N = 369, 63.00%). Most participants reported a monthly income of 3,001–5,000 (N = 190, 32.40%). The sample’s mean age was 35.14 ± 13.53 years. Detailed sociodemographic data are presented in Table 1.

Demographic information of the participants.

Variables Items Number Proportion
Gender Female 225 38.40%
Male 361 61.60%
Educational background Primary school and below 162 27.60%
Junior high school to Senior high school 245 41.80%
Bachelor’s degree or above 179 30.50%
Marriage Unmarried 158 27.00%
Married 336 57.30%
Divorced 73 12.50%
Widowed 19 3.20%
Place of residence Urban 369 63.00%
Rural areas 217 37.00%
Monthly income ≤1,000 61 10.40%
1,001–3,000 140 23.90%
3,001–5,000 190 32.40%
5,001–8,000 156 26.60%
≥8,001 39 6.70%
Causes of trauma Traffic accident 167 28.50%
Tumble 130 22.20%
Violent attack 121 20.60%
Sport injury 168 28.70%
Trauma staging Acute stage 132 22.50%
Subacute stage 211 36.00%
Chronic stage 243 41.50%
Measures tools Self-compassion scale

Self-compassion was assessed using the Short Form Self-compassion Scale developed by Raes, Pommier (63), consisting of 26 items across 6 dimensions: self-kindness (5 items), self-criticism (5 items), common humanity (4 items), isolation (4 items), mindfulness (4 items), and over-identification (4 items). The scale was adapted and validated for Chinese adolescents (Chinese version), demonstrating cultural applicability, reliability, and validity (64). It has also been widely applied in patient populations such as breast cancer (65), major depressive disorder (66), and brain injury (67, 68). Responses were rated on a 5-point Likert scale (1 = “almost never” to 5 = “almost always”). Positive subscales (self-kindness, common humanity, mindfulness) and negative subscales (self-criticism, isolation, over-identification) jointly generated an overall self-compassion score. Total scores range from 26 to 130, with higher scores indicating greater self-compassion. Confirmatory factor analysis (CFA) was performed in AMOS 29.0, and reliability testing in SPSS 27.0, showing good construct validity and internal consistency (χ2/df = 2.647, CFI = 0.904, GFI = 0.896, AGFI = 0.873, TLI = 0.891, RMSEA = 0.053; Cronbach’s α = 0.921).

 Self-worth scale

Self-worth was measured with the Contingencies of Self-worth Scale developed by Crocker, Luhtanen (69), consisting of 35 items across 7 domains: family support (5 items), competition (5 items), appearance (5 items), God’s love (5 items), academic competence (5 items), virtue (5 items), and others’ approval (5 items). The Chinese version was adapted by Cheng and Kwan (70) and later validated by Geng and Jiang (71). This scale has also been applied in populations such as cosmetic surgery patients (72), obsessive-compulsive disorder (73), and multiple sclerosis (74). Responses were rated on a 5-point Likert scale (1 = “strongly disagree” to 5 = “strongly agree”), with total scores ranging from 35 to 175. Higher scores indicate greater perceived self-worth. CFA and reliability testing confirmed strong psychometric properties (χ2/df = 2.632, CFI = 0.887, GFI = 0.845, AGFI = 0.825, TLI = 0.880, RMSEA = 0.053; Cronbach’s α = 0.950).

 Social support scale

Perceived social support was assessed using the Multidimensional Scale of Perceived Social Support (MSPSS) developed by Zimet, Dahlem (75). The MSPSS includes 12 items across three subscales: significant others (4 items), family (4 items), and friends (4 items). The Chinese version was validated in university populations (76) and has demonstrated strong reliability in brain injury populations (77, 78). Responses were rated on a 5-point Likert scale (1 = “strongly disagree” to 5 = “strongly agree”). A mean score across all items was calculated, with possible scores ranging from 1 to 5; higher scores reflect stronger perceived social support. In the subsequent multinomial logistic regression analysis, Odds Ratios for social support represent the change in odds associated with each 1-point increase on this 5-point scale. Good model fit and reliability were confirmed (χ2/df = 2.416, CFI = 0.962, GFI = 0.964, AGFI = 0.947, TLI = 0.953, RMSEA = 0.049; Cronbach’s α = 0.870).

 Death anxiety scale

Death anxiety was measured by the Death Anxiety Scale (DAS) originally developed by Templer (79). It consists of 15 items assessing anxiety responses to death, illness, surgery, time, war, and exposure to corpses. The Chinese version was adapted and validated in colorectal cancer patients (80). The scale has been widely applied in cancer (81), chronic illness (82), and chronic back pain (83). Participants rated items on a 5-point Likert scale (1 = “strongly disagree” to 5 = “strongly agree”). A mean score across all items was calculated, with possible scores ranging from 1 to 5; higher scores indicate greater death anxiety. In the subsequent multinomial logistic regression analysis, Odds Ratios for death anxiety represent the change in odds associated with each 1-point increase on this 5-point scale. CFA and reliability analyses supported adequate construct validity and internal consistency (χ2/df = 2.279, CFI = 0.937, GFI = 0.954, AGFI = 0.939, TLI = 0.926, RMSEA = 0.047; Cronbach’s α = 0.822).

 Statistical analyses

All statistical analyses were conducted using SPSS 27.0, AMOS 29.0, and Mplus 8. First, CFA was performed in AMOS 29.0 to evaluate model fit for the core measures. Second, SPSS 27.0 was employed for common method bias testing, descriptive statistics, and correlation analyses. Third, Mplus 8 was used to construct a five-profile latent profile model of self-compassion and self-worth; selection of the optimal model was based on Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), Adjusted BIC (aBIC), and Entropy. Subgroup profiles for the optimal model were then generated. In LPA, individuals are probabilistically assigned to the latent profile with the highest posterior probability; entropy was used to evaluate classification precision. Fourth, univariate tests assessed differences in demographics, social support, and death anxiety across self-compassion - self-worth latent profiles. Finally, multivariate logistic regression tested the predictive value of significant variables from univariate analyses.

Model fit indices included: Comparative Fit Index (CFI ≥ 0.80 acceptable), Goodness-of-Fit Index (GFI ≥ 0.80), Adjusted Goodness-of-Fit Index (AGFI ≥ 0.80), Tucker-Lewis Index (TLI ≥ 0.80), and Root Mean Square Error of Approximation (RMSEA ≤ 0.08). Profile fit criteria included smaller values of AIC, BIC, and aBIC, while Entropy > 0.80 indicated excellent classification quality. A p-value < 0.05 was considered statistically significant.

 Result Common method bias analysis

To further reduce common method bias, this study employed anonymization and confidentiality during the data collection phase to mitigate participants’ social desirability tendencies and evaluation concerns. Subsequently, we utilized Harman’s single-factor test, conducting exploratory factor analysis on all variable measurement items to examine the unrotated factor solution. The results indicated that the first factor explained 23.94% of the variance, which did not exceed the threshold of 40%, suggesting that this study does not suffer from serious common method bias.

 Descriptive statistics and correlation analysis

We conducted descriptive statistics and correlation analysis on demographic information, self-compassion, self-worth, social support, and death anxiety, as presented in Tables 2, 3. The levels of self-compassion (M = 3.142, SD = 0.657), self-worth (M = 3.121, SD = 0.685), social support (M = 3.015, SD = 0.736), and death anxiety (M = 3.293, SD = 0.679) were all above the midpoint (M = 2.5). Meanwhile, the skewness of the core variables ranged from −0.787 to 0.061, and kurtosis ranged from −0.193 to −0.458. According to Kline (84) findings, the core variables in this study exhibited an approximately normal distribution.

Descriptive statistical analysis of core variables.

Variables M SD Skewness Kurtosis
Self-compassion 3.142 0.657 0.061 −0.127
Self-worth 3.121 0.685 −0.204 −0.193
Social support 3.015 0.736 −0.163 −0.100
Death anxiety 3.293 0.679 −0.787 0.458

Correlation analysis of demographic information and core variables.

Variable 1 2 3 4 5 6 7 8 9 10 11 12

Self-compassion

 		1

Self-worth

 		0.571** 1
3.Social support 0.306** 0.622** 1
4.Death anxiety −0.480** −0.526** −0.144** 1
5.Gender −0.242** −0.367** −0.146** 0.265** 1

Age

 		−0.002 0.008 0.002 0.012 −0.012 1

Education background

 		−0.116** −0.185** −0.144** 0.102* 0.016 −0.406** 1

Marriage

 		0.044 0.058 0.046 −0.057 0.014 0.680** −0.263** 1

Place of residence

 		0.067 −0.003 −0.062 −0.044 0.024 0.001 −0.001 0.110** 1

Monthly income

 		0.039 0.03 0.001 0.028 0.004 −0.122** −0.076 −0.014 0.011 1

Causes of trauma

 		0.005 −0.100* −0.098* 0.04 0.06 −0.102* 0.100* −0.100* −0.094* −0.033 1

Trauma staging

 		−0.013 0.02 0.075 0.02 −0.02 −0.029 −0.018 −0.031 −0.005 0.015 −0.046 1

*p < 0.05; **p < 0.01.

Table 3 displays the complete correlation matrix. The study found that self-compassion had a significant positive correlation with self-worth (r = 0.571, p < 0.01), a significant positive correlation with social support (r = 0.306, p < 0.01), and a significant negative correlation with death anxiety (r = −0.408, p < 0.01). Self-worth also showed a significant positive correlation with social support (r = 0.626, p < 0.01) and a significant negative correlation with death anxiety (r = −0.526, p < 0.01). Social support was significantly negatively correlated with death anxiety (r = −0.144, p < 0.01).

Regarding demographic information, education and gender significantly correlated with self-compassion, self-worth, social support, and death anxiety (p < 0.01). The cause of trauma significantly correlated with self-worth and social support (p < 0.05). However, age, marital status, residence, monthly income, and trauma phase showed no significant correlations with the core variables (p > 0.05).

 Latent profile analysis of self-compassion and self-worth

Using the dimension scores of self-compassion and self-worth as indicators, we employed Mplus 8 software to construct a latent profile model for TBI patients, categorizing them into 1 to 5 classes. As shown in Table 4, with an increasing number of latent profiles, the values of AIC, BIC, and aBIC consistently decreased, indicating an improvement in the fit of the latent profile model. Compared to the two-class latent profile model (Entropy = 0.934), the three-class latent profile model exhibited superior classification clarity, providing more reliable results (Entropy = 0.946). To determine whether adding an additional profile significantly improved model fit, we used the Lo–Mendell–Rubin adjusted likelihood ratio test (LMR-LRT), which compares a k-profile model against a (k–1)-profile model. The LMR-LRT was significant for the three-profile model (p = 0.016; 3 vs. 2 profiles), suggesting that the three-profile solution fit significantly better than the two-profile solution. In contrast, the LMR-LRT was not significant for the four-profile model (p = 0.172; 4 vs. 3 profiles), indicating that adding a fourth profile did not provide a statistically significant improvement over the three-profile solution. Considering the nonsignificant incremental gain for the 4-profile model, the higher entropy for the three-profile solution (0.946), and the principle of parsimony and interpretability, we selected the three-profile model as the optimal solution.

One to five fitting indicators of potential profile models.

Profile AIC BIC aBIC Entropy LMR(P) BLRT(P) Smallest proportion per class
1 18299.824 18413.530 18330.990
2 15535.798 15710.731 15583.745 0.934 <0.001 <0.001 0.653/0.347
3 14368.002 14604.162 14432.731 0.946 0.016 <0.001 0.188/0.258/0.554
4 13485.930 13783.316 13567.440 0.936 0.172 <0.001 0.231/0.467/0.066/0.236
5 13140.906 13499.518 13239.197 0.914 0.139 <0.001 0.136/0.407/0.066/0.224/0.167

To evaluate the robustness of the three-class solution, particularly given that Class 1 comprised 18.8% of the sample, we conducted several sensitivity analyses. While this proportion approaches commonly discussed thresholds for small class sizes in LPA, it substantially exceeds the recommended minimum criteria. Methodological guidelines suggest caution for classes containing fewer than 5% of the sample or fewer than 25 individuals (85, 86). Class 1, with 110 participants, far exceeds these thresholds. We performed split-sample cross-validation by randomly dividing the sample into two halves. Independent LPA on each subsample consistently yielded a three-class solution, with Class 1 proportions of 17.4 and 19.8% respectively, demonstrating stability across random subsamples.

Based on the analysis results in Table 4, we utilized Origin 2021 software to plot the profile graph of the three-class latent profile model, as shown in Figure 1. Figure 1 illustrates that the first latent subgroup displays solid recognition and positive evaluation of self-compassion and self-worth, labeled as “High self-compassion – high self-worth,” comprising 25.8% of the participants. The second latent subgroup exhibits moderate recognition and evaluation of self-compassion and self-worth, named “Moderate self-compassion - moderate self-worth,” making up 55.4%. The third latent subgroup represents patients with moderate self-compassion and lower self-worth, identified as “Moderate self-compassion – low self-worth,” accounting for 18.8%.

Latent profile plot of standardized mean scores (z-scores) across the measured dimensions of self-compassion and self-worth. The x-axis lists each dimension in the order used in the latent profile analysis: Self-kindness, Self-criticism, Common humanity, Isolation, Mindfulness, Over-identification, Family support, Competition, Appearance, God’s love, Academic competence, Virtue, Others’ approval. Higher values indicate higher levels on the corresponding dimension.

Line graph showing mean values across variable dimensions for three groups: moderate self-compassion and low self-worth (gray), high self-compassion and high self-worth (red), and moderate self-compassion and moderate self-worth (blue). High self-compassion group consistently scores highest across dimensions. Univariate analysis of latent subgroups

Before conducting univariate comparisons across subgroups, we first determined the optimal number of latent profiles. We fitted a series of LPA models with increasing numbers of profiles using self-compassion and self-worth indicators. Model fit improved from 2 to 3 profiles, as reflected by lower information criteria and a significant LMR-LRT for the 3-profile model (p = 0.016; 3 vs. 2 profiles). In contrast, the LMR-LRT for the 4-profile model was not significant (p = 0.172; 4 vs. 3 profiles), indicating that adding a fourth profile did not yield a statistically meaningful improvement over the 3-profile solution. The 3-profile solution also showed high classification accuracy (entropy = 0.946) and yielded interpretable profiles with adequate sample sizes. Therefore, we selected the 3-profile model for subsequent analyses.

After selecting the three-profile solution, we compared demographic, injury-related, and psychosocial variables across the latent classes using univariate tests (χ2 tests for categorical variables; ANOVA or nonparametric tests as appropriate for continuous variables). Variables showing significant between-class differences were then entered into a multinomial logistic regression model to identify independent correlates of class membership.

Significant gender differences were observed (χ2 = 75.138, p < 0.001), with a higher proportion of women in Class 1 (14.4%) and Class 3 (40.1%), while Class 2 had an even higher female proportion (73.0%), indicating that women are more likely to fall into the lower self-worth profile. Educational level differences were significant (χ2 = 21.774, p < 0.001), with a higher proportion of individuals in Class 2 holding a bachelor’s degree or higher (24.1%), while Classes 1 and 3 had a greater proportion of individuals with lower educational levels. There were no significant differences in marital status (χ2 = 7.258, p = 0.298) and monthly income (χ2 = 10.525, p = 0.230), suggesting these factors do not distinguish the profiles.

Differences in residence were marginally significant (χ2 = 6.471, p = 0.039), with a slightly higher proportion of urban residents in Class 3 (77.8%), while Classes 1 (61.9%) and 2 (66.2%) had relatively lower proportions. The cause of injury was significantly associated with profile membership (χ2 = 18.157, p = 0.006), with traffic accidents being more prevalent in Class 3 (36.9%), whereas sports injuries dominated Class 2 (30.8%). Differences in injury phases were significant (χ2 = 12.533, p = 0.014), with chronic phase patients over represented in Class 3 (51.1%) and acute phase patients more prevalent in Class 2 (24.1%).

There were no significant age differences among profiles (F = 0.184, p = 0.832). However, perceived social support showed significant differences (F = 173.645, p < 0.001), with Class 2 having the highest level (M = 3.527, SD = 0.789), followed by Class 3 (M = 3.061, SD = 0.736), and Class 1 the lowest (M = 2.164, SD = 0.549). A one-way ANOVA confirmed all pairwise comparisons were significant (p < 0.001). Death anxiety exhibited a similar pattern (F = 267.243, p < 0.001): Class 2 scored the lowest (M = 2.494, SD = 0.617), Class 1 was moderate (M = 3.522, SD = 0.446), and Class 3 was the highest (M = 3.584, SD = 0.439), with all pairwise comparisons being significant (p < 0.001).

These univariate results emphasize that the profile with higher self-compassion and self-worth (Class 2) is associated with greater social support, lower death anxiety, higher education levels, and specific injury characteristics, while the lower profile (Class 1) shows the opposite trend (Table 5).

Univariate analysis of potential subgroups.

Variable Items Class 1 Class 2 Class 3 χ2/F p
Gender Female 14 97 114 75.138 <0.001
Male 94 53 214
Educational background Primary school and below 24 57 81 21.774 <0.001
Junior high school to Senior high school 36 61 148
Bachelor’s degree or above 48 32 99
Marriage Unmarried 27 31 100 7.258 0.298
Married 64 90 182
Divorced 15 23 35
Widowed 2 6 11
Place of residence Urban 60 88 221 6.471 0.039
Rural areas 48 62 107
Monthly income ≤1,000 9 16 36 10.525 0.230
1,001–3,000 35 32 73
3,001–5,000 39 47 104
5,001–8,000 19 42 95
≥8,001 6 13 20
Causes of trauma Traffic accident 18 44 105 18.157 0.006
Tumble 25 42 63
Violent attack 23 23 75
Sport injury 42 41 85
Trauma staging Acute stage 36 32 64 12.533 0.014
Subacute stage 29 63 119
Chronic stage 43 55 145
Age 35.58 ± 12.81 34.61 ± 12.59 35.24 ± 14.19 0.184 0.832
Social support 2.164 ± 0.549 3.527 ± 0.789 3.061 ± 0.736 173.645 <0.001
Death anxiety 3.522 ± 0.446 2.494 ± 0.617 3.584 ± 0.439 267.243 <0.001

Class 1: Moderate self-compassion - low self-worth group; Class 2: High self-compassion - high self-worth group; Class 3: Moderate self-compassion - moderate self-worth group.

 Multivariate logistic regression analysis of latent subgroups

To identify independent predictors of profile membership, we performed multivariate logistic regression using the significant variables from the univariate analysis (p < 0.05: gender, education, residence, cause of injury, injury phase, social support, death anxiety). Class 3 (“Moderate self-compassion - moderate self-worth”) was used as the reference category due to its maximum sample size and intermediate positioning. Two models were estimated, comparing Class 1 to Class 3 and Class 2 to Class 3, with results presented in Table 6. Notably, for continuous variables (social support and death anxiety), scores were calculated as mean item scores on 5-point Likert scales (range: 1–5), and the reported Odds Ratios represent the change in odds for each 1-point increase on these scales.

Multiple regression analysis (with class 3 as the reference group).

Potential subgroups Variables Dimensionality Regression Coefficient Standard Error Wald χ2 P OR LLCI ULCI
Class 1 Social support −3.009 0.337 79.628 <0.001 0.049 0.025 0.096
Death anxiety 0.331 0.351 0.889 0.346 1.392 0.700 2.771
Gender Female −1.3 0.392 11.03 0.001 0.272 0.126 0.587
Male (refer)
Educational background Primary school and below −0.464 0.398 1.356 0.244 0.629 0.288 1.373
Junior high school to Senior high school −0.349 0.341 1.045 0.307 0.705 0.361 1.377
Bachelor’s degree or above (refer)
Place of residence Urban −0.107 0.317 0.113 0.736 0.899 0.483 1.672
Rural areas (refer)
Causes of trauma Traffic accident −0.649 0.408 2.532 0.112 0.522 0.235 1.162
Tumble −0.631 0.426 2.200 0.138 0.532 0.231 1.225
Violent attack −0.763 0.415 3.375 0.066 0.466 0.207 1.052
Sport injury (refer)
Trauma staging Acute stage 0.228 0.367 0.383 0.536 1.256 0.611 2.58
Subacute stage −0.452 0.373 1.466 0.226 0.636 0.306 1.323
Chronic stage (refer)
Class 2 Social support 1.914 0.303 39.872 <0.001 6.782 3.744 12.287
Death anxiety −3.938 0.395 99.555 <0.001 0.019 0.009 0.042
Gender Female 1.249 0.365 11.676 0.001 3.485 1.703 7.132
Male (refer)
Educational background Primary school and below 1.133 0.504 5.054 0.025 3.104 1.156 8.335
Junior high school to Senior high school 0.616 0.466 1.745 0.187 1.851 0.742 4.614
Bachelor’s degree or above (refer)
Place of residence Urban −0.342 0.377 0.825 0.364 0.71 0.34 1.486
Rural areas (refer)
Causes of trauma Traffic accident −0.444 0.479 0.859 0.354 0.642 0.251 1.64
Tumble 0.262 0.479 0.299 0.584 1.3 0.508 3.326
Violent attack −1.045 0.559 3.490 0.062 0.352 0.118 1.053
Sport injury (refer)
Trauma staging Acute stage 0.656 0.473 1.923 0.165 1.927 0.762 4.872
Subacute stage 0.236 0.407 0.336 0.562 1.266 0.57 2.808
Chronic stage (refer)

Class 1: Moderate self-compassion - low self-worth group; Class 2: High self-compassion - high self-worth group.

For the comparison of Class 1 (“Moderate self-compassion - low self-worth”) to Class 3: social support emerged as a significant protective factor (β = −3.009, SE = 0.337, Wald χ2 = 79.628, p < 0.001, OR = 0.049, 95% CI = [0.025, 0.096]), indicating that higher social support significantly reduces the likelihood of falling into this low self-worth profile. This indicates that for every 1-point increase in social support on the 5-point scale, the odds of belonging to the low self-worth profile (Class 1) versus the moderate profile (Class 3) decrease by approximately 95% (OR = 0.049). Given that the sample mean for social support was 3.015 (SD = 0.736), a 1-point increase represents a substantial shift of approximately 1.4 standard deviations. Death anxiety was not significant (β = 0.331, SE = 0.351, Wald χ2 = 0.889, p = 0.346, OR = 1.392, 95% CI = [0.700, 2.771]). Gender was a key predictive factor, with women less likely to fall into Class 1 (β = −1.300, SE = 0.392, Wald χ2 = 11.030, p = 0.001, OR = 0.272, 95% CI = [0.126, 0.587]); specifically, women’s odds of falling into the low self-worth profile were approximately 73% lower than men’s. Education, residence, cause of injury, and injury phase did not reach significance (all p > 0.05), indicating that these effects were weakened when controlling for other variables.

For the comparison of Class 2 (“High self-compassion - high self-worth”) to Class 3: social support was a strong positive predictor (β = 1.914, SE = 0.303, Wald χ2 = 39.872, p < 0.001, OR = 6.782, 95% CI [3.744, 12.287]), with higher levels increasing the likelihood of falling into the high profile. This indicates that for every 1-point increase in social support on the 5-point scale, the odds of belonging to the high profile (Class 2) versus the moderate profile (Class 3) increase by a factor of approximately 6.8. Death anxiety served as a protective factor (β = −3.938, SE = 0.395, Wald χ2 = 99.555, p < 0.001, OR = 0.019, 95% CI = [0.009, 0.042]); for every 1-point increase in death anxiety, the odds of belonging to the high profile decrease by approximately 98% (OR = 0.019), underscoring the powerful negative association between existential anxiety and optimal self-perception profiles. Gender was significant, with women having approximately 3.5 times higher odds of belonging to Class 2 compared to men (β = 1.249, SE = 0.365, Wald χ2 = 11.676, p = 0.001, OR = 3.485, 95% CI = [1.703, 7.132]). Education showed a significant effect for the lowest category: individuals with primary school education or below had approximately 3.1 times higher odds of belonging to the high profile compared to those with bachelor’s degree or above (β = 1.133, SE = 0.504, Wald χ2 = 5.054, p = 0.025, OR = 3.104, 95% CI = [1.156, 8.335]), while the middle education category (junior high to senior high school) was not significant (p = 0.187). Residence, cause of injury, and injury phase were not significant (all p > 0.05).

Overall, these multivariate results underscore social support and death anxiety as robust psychosocial predictors of latent profile membership, with gender and education exerting additional independent effects. Demographic and injury-related factors appeared less critical when controlling for psychosocial variables, highlighting potential intervention targets in the rehabilitation of TBI patients.

 Discussion Latent profile analysis of self-compassion and self-worth

This study identified three heterogeneous subtypes of the relationship between self-compassion and self-worth in patients with TBI through LPA: the High self-compassion - high self-worth group (Class 2, 25.8%), the Moderate self-compassion - moderate self-worth group (Class 3, 55.4%), and the Moderate self-compassion - low self-worth group (Class 1, 18.8%). This finding reveals significant internal heterogeneity in self-perception within this patient population, contrasting with previous variable-centered research methods that often assume homogeneity and overlook individual differences. Specifically, patients in the High self-compassion - high self-worth group demonstrated elevated levels of both self-compassion and self-worth, potentially reflecting their stronger abilities for self-acceptance and positive self-evaluation mechanisms, thereby buffering against the negative impacts of trauma. In contrast, the Moderate self-compassion - moderate self-worth group, as the largest subtype, exhibited a balanced level of self-perception, indicating that most patients maintained relatively stable self-regulation strategies post-trauma. However, patients in the Moderate self-compassion – low self-worth group displayed moderate self-compassion but significantly lower self-worth, highlighting a potential disconnect where self-compassion did not effectively translate into self-worth for some individuals, possibly stemming from identity crises and loss of social roles triggered by trauma.

These profile patterns partially align with findings in existing literature on chronic illness and post-traumatic groups. For example, similar LPA studies in breast cancer patients have identified high, moderate, and low self-compassion subtypes (87), confirming that the high self-compassion group is associated with better psychological adjustment. However, this study uniquely incorporates both self-compassion and self-worth in the LPA indicators, extending the limitations of previous research that focused solely on single variables. The imbalanced pattern observed in the Moderate self-compassion - low self-worth group may arise from specific neurocognitive impairments associated with TBI, such as difficulties in reflection and emotional regulation due to frontal lobe dysfunction, which hinder the transformation of self-compassion into self-worth (88). This heterogeneity emphasizes the nonlinear characteristics of psychological recovery post-trauma: while self-compassion can serve as a protective factor, its effectiveness depends on individual backgrounds and trauma severity (16). Additionally, the proportion of low self-worth subtypes among TBI patients is relatively low, which may be attributed to the longer duration of illness in this sample (predominantly in the chronic phase), allowing some patients to reconstruct self-perception through the rehabilitation process.

From a theoretical perspective, these profiles support the conditional self-worth theory, wherein self-worth is influenced by achievements and social recognition. TBI often leads to functional loss and social isolation, resulting in fluctuations in self-worth (24). Meanwhile, self-compassion, as a buffering mechanism, may strengthen the stability of self-worth in the high-profile group through components of self-kindness and mindfulness (89). However, the disconnect in the moderate to low profile group suggests that simply enhancing self-compassion may be insufficient to restore self-worth, especially in the presence of death anxiety or insufficient social support. This finding provides insights for clinical practice: personalized interventions should be tailored to different profiles, such as emphasizing self-worth reconstruction training (e.g., cognitive behavioral therapy combined with self-compassion meditation) for the moderate to low group, while focusing on maintenance strategies for the high group to prevent potential declines. In summary, the results of the LPA in this study not only enrich the empirical foundation of psychological heterogeneity in TBI but also lay a theoretical framework for precise psychological interventions.

 Analysis of influencing factors for latent subgroups

This study revealed key factors influencing the latent profile membership of self-compassion and self-worth in patients with TBI through multivariate logistic regression analysis, including social support, death anxiety, gender, and education level. Residence, cause of injury, and injury phase did not show independent predictive effects after controlling for other variables. This finding emphasizes the central role of psychosocial factors in the heterogeneity of self-perception post-trauma, resonating with social support buffering models and fear management theories. Specifically, social support emerged as a strong protective factor, positively correlating with the High self-compassion - high self-worth group and negatively correlating with the Moderate self-compassion - low self-worth group. This indicates that adequate social support can enhance patients’ self-kindness and stability of self-worth through emotional validation and resource provision (90). This result aligns with existing literature on TBI patients, where perceived social support significantly reduces psychological fatigue and feelings of isolation, thereby promoting positive self-perception reconstruction (91). Further research has shown that social support plays a crucial role in recovery from TBI, improving emotional regulation and buffering against self-criticism triggered by trauma (92, 93). However, the average level of social support in this study’s sample was relatively high, which may have amplified its protective effects; future studies should validate its threshold in low-support groups to optimize intervention strategies.

Death anxiety also served as another independent predictive factor, negatively correlating with the High self-compassion - high self-worth group, suggesting that lower death anxiety helps maintain a positive self-perception profile, while higher anxiety may disrupt the buffering mechanisms of self-compassion, leading to a decline in self-worth. This association supports the view of death anxiety as a cross-diagnostic construct, often amplifying negative self-evaluations in trauma patients by activating existential threats. In the context of TBI patients, death anxiety may arise from awareness of functional loss and life fragility, closely linked to low self-worth and depressive symptoms (94). Although death anxiety did not reach significant independent effects in the low-profile group, its overall negative correlation indicates it may influence profile membership through indirect pathways. The moderate level of death anxiety among TBI patients may reflect the abruptness of trauma rather than a gradual process, yet it still emphasizes the need to integrate existential therapies (e.g., meaning-oriented interventions) to reduce anxiety and enhance self-worth.

Gender differences were significantly evident in this study, with females more likely to belong to the High self-compassion - high self-worth group, while males were more prone to the low profile group. This may stem from women having higher baseline levels of self-compassion, including stronger components of self-kindness and mindfulness (95, 96). Existing literature confirms that women often report higher scores on self-compassion scales, although they may also exhibit higher scores on non-compassion dimensions (e.g., self-criticism) (95, 97). The gender differences in TBI patients may further amplify: women excel in emotional empathy and self-concept reconstruction, while men may suppress expressions of self-compassion due to social norms (e.g., rigid masculinity), leading to a greater vulnerability in self-worth. This finding aligns with gender differences in overall outcomes for TBI, where women face more challenges in physiological recovery but demonstrate stronger psychological adaptability.

Education level emerged as an additional predictive factor, with the low education group (elementary school or below) more likely to belong to the High self-compassion - high self-worth group. This counterintuitive result may indicate that individuals with lower education levels rely more on intrinsic sources of self-worth (e.g., familial support or personal resilience) rather than external achievements (e.g., occupational recognition), thereby buffering the impact of trauma on self-perception. Literature indicates that education level positively correlates with self-efficacy and quality of life in chronic disease patients, with higher educated individuals often demonstrating better self-care (98, 99). However, lower educated individuals might develop stronger internal compassion through simplified cognitive frameworks. In chronic illness populations, low education is associated with higher depression; yet, the association of high profiles in this study suggests that education may influence self-worth through health literacy. This finding challenges traditional assumptions and emphasizes the need to consider the interplay of education and cultural context: in low-income areas, lower educated patients may benefit more from community support. Other factors, such as residence, cause of injury, and injury phase, were not significant in the multivariate analysis, suggesting their effects may be mediated by psychosocial variables (e.g., social support absorbing urban–rural differences), which aligns with the limitations of variable-centered studies.

 Education level interpretation

Education level emerged as an additional predictive factor with a seemingly paradoxical pattern: individuals with the lowest education level (primary school or below) demonstrated approximately 3.1 times higher odds of belonging to the High self-compassion - high self-worth group compared to those with bachelor’s degree or above. This counterintuitive finding warrants careful interpretation within the Chinese cultural context and consideration of multiple explanatory mechanisms.

First, this pattern may reflect the influence of collectivist cultural values predominant in Chinese society, particularly among older and less formally educated populations. In collectivist cultures, self-worth is often more strongly anchored in relational and communal domains rather than individual achievements (100, 101). For individuals with lower formal education, self-worth may be primarily derived from family roles, community standing, and moral character rather than educational or occupational accomplishments. Consequently, the functional impairments and occupational disruptions caused by TBI may pose less threat to their self-worth foundations compared to highly educated individuals whose self-worth is more contingent upon achievement and professional identity.

Second, traditional Chinese philosophical frameworks, particularly Confucian concepts of self-cultivation and Buddhist teachings on acceptance and non-attachment, may provide cognitive resources that promote self-compassion independent of educational attainment (102). These philosophical traditions, often more deeply embedded in rural and less-educated communities through oral transmission and cultural practices, emphasize inner contentment, acceptance of life circumstances, and compassionate self-treatment during hardship. Such cultural resources may enable individuals with lower formal education to maintain self-compassion and self-worth despite traumatic experiences.

Third, structural differences in social support networks may partially explain this finding. In Chinese society, individuals with lower education often reside in rural areas or maintain closer-knit family and community networks characterized by stronger reciprocal support obligations (103). These dense social networks may provide more robust emotional validation and practical assistance following TBI, thereby bolstering both self-compassion and self-worth. Although we controlled for perceived social support in our regression model, unmeasured qualitative aspects of support—such as cultural appropriateness, integration into daily life, and alignment with self-worth sources—may differ systematically across education levels.

 Practical implications

The findings of this study provide important practical guidance for clinical interventions in patients with TBI by revealing the heterogeneous profiles of the relationship between self-compassion and self-worth and their influencing factors. This underscores the necessity of adopting a person-centered approach to develop precise psychological rehabilitation strategies. Specifically, the identified three latent profiles can serve as a foundation for clinical screening, assisting medical teams in quickly categorizing patients and matching targeted interventions, thereby optimizing resource allocation and enhancing treatment efficacy.

Profile-specific intervention recommendations

Class 1: Moderate Self-Compassion – Low Self-Worth Group. This profile represents the highest-risk subgroup, characterized by a disconnect where self-compassion fails to translate into stable self-worth. For these patients, we recommend a multi-component intervention approach. First, Compassion-Focused Therapy (CFT) should be prioritized, specifically targeting the self-worth reconstruction pathway. CFT’s emphasis on developing the “soothing system” through compassionate imagery and self-reassurance exercises can address the self-criticism that may underlie low self-worth despite moderate self-compassion. Prior research by Ownsworth and Haslam (23) demonstrated that integrating CFT with cognitive retraining produced significant improvements in self-evaluation among brain injury patients. Second, Cognitive-Behavioral Therapy (CBT) modules focusing on contingent self-worth restructuring should be incorporated. Given that this profile shows vulnerability in self-worth contingencies, therapeutic work should help patients identify and challenge maladaptive self-worth bases (e.g., appearance, occupational achievement) that have been disrupted by injury, while cultivating intrinsic sources of self-worth (e.g., personal values, relationships). Third, social support enhancement interventions are critical, as this group demonstrated the lowest social support levels (M = 2.164). Concrete approaches include family psychoeducation sessions to improve caregivers’ capacity for emotional validation, peer support group participation connecting patients with successfully rehabilitated individuals, and community resource linkage to address instrumental support gaps.

Class 2: High Self-Compassion – High Self-Worth Group. This profile represents patients with optimal psychological adaptation who require maintenance-focused rather than intensive interventions. Recommended approaches include preventive mindfulness-based relapse prevention to sustain high self-compassion levels, regular psychological monitoring (quarterly assessments) to detect early signs of decline, and resilience-building psychoeducation empowering patients as peer mentors for lower-profile patients. These patients may also serve as valuable resources in group interventions, modeling adaptive self-perception strategies for other participants.

Class 3: Moderate Self-Compassion – Moderate Self-Worth Group. As the largest subgroup with balanced but potentially fragile self-perception, these patients benefit from standard supportive interventions with targeted enhancement: Brief self-compassion training using Neff’s Mindful Self-Compassion (MSC) protocol adapted for TBI populations, focusing on the self-kindness and common humanity components; psychoeducation addressing post-traumatic identity adjustment and normalizing psychological challenges following TBI; and booster sessions during high-risk periods (e.g., rehabilitation setbacks, anniversary reactions) to prevent deterioration into lower profiles.

 Practice implications of social support and death anxiety

The discovery of social support as a key protective factor suggests that clinicians should incorporate social network reconstruction into rehabilitation plans. For instance, providing emotional and instrumental resources through family involvement groups or community support programs can significantly enhance patients’ self-acceptance and buffer against self-worth crises, consistent with empirical evidence of social support in brain injury recovery. In practical terms, mobile application-based virtual support networks could be developed to provide remote interventions for patients in urban–rural disparity, overcoming geographical barriers and enhancing accessibility. Simultaneously, the negative predictive role of death anxiety emphasizes the need to integrate existential or meaning-oriented therapies, such as managing death anxiety modules, to help patients reconstruct life narratives and reduce existential threats, thereby indirectly enhancing self-worth. This strategy is particularly suitable for the high-anxiety subgroup and can be combined with self-compassion induction exercises to reduce the activation of the threat response system and improve overall psychological adaptation. Given the significant predictive role of death anxiety, particularly for Class 2 membership, we recommend integrating existential therapy components across all profiles. Meaning-centered interventions, adapted from Breitbart’s work with cancer patients, can help TBI patients reconstruct life narratives incorporating the trauma experience. For patients with elevated death anxiety, specific modules addressing mortality acceptance—such as life review exercises and values clarification—can reduce existential distress and indirectly support self-worth stability.

 Gender-responsive intervention design

The finding that women were more likely to belong to the high self-compassion–high self-worth profile, while men were overrepresented in lower profiles, has important implications for gender-responsive care. Interventions for male TBI patients should explicitly address barriers to self-compassion, including rigid masculinity norms that discourage emotional vulnerability and self-kindness. Group interventions segregated by gender may facilitate more open discussion of these barriers, and psychoeducation normalizing self-compassion as a strength rather than weakness may improve engagement.

The independent effects of gender and education level further highlight the need for personalized interventions. For male patients, modules should be designed to challenge rigid self-norms and promote self-compassion expression to reduce the risk of falling into low self-worth profiles; for those with lower education, simplified, visually aided intervention forms can leverage their internal resilience advantages to strengthen positive self-perception. On a broader scale, this study supports the development of clinical assessment tools based on LPA, such as a digital screening system integrating self-compassion and self-worth scales for early identification of at-risk subtypes and guiding multidisciplinary teams (e.g., neurosurgeons, psychotherapists, and social workers) in collaborative interventions. Meta-analytic evidence indicates that self-compassion interventions overall can improve psychosocial outcomes, but the unique neurobiological foundations of brain injury must be considered to avoid limited efficacy in severe cases, such as compassion imagery interventions. Additionally, at the policy level, these findings could drive updates in public health guidelines, such as incorporating self-compassion training modules into TBI rehabilitation protocols in China’s tiered healthcare system, to enhance patient quality of life and reduce healthcare burdens.

 Implementation considerations

At the systems level, these findings support the development of a stepped-care model for TBI psychological rehabilitation. Initial screening using abbreviated self-compassion and self-worth measures could efficiently stratify patients into risk categories, with intervention intensity matched to profile membership. A digital screening tool integrating these assessments could be embedded in electronic health records, prompting automated referrals to appropriate intervention levels.

For resource-constrained settings, particularly in low- and middle-income countries where TBI prevalence is highest, scalable interventions such as mobile application–delivered self-compassion exercises or trained lay counselor–facilitated support groups offer pragmatic alternatives to specialist-delivered therapy. The social support finding specifically supports investment in community health worker programs to extend psychological support reach.

 Limitations and future research directions

Despite providing empirical evidence for the heterogeneity of the relationship between self-compassion and self-worth in patients with TBI, this study has several limitations that require cautious interpretation of the results and guidance for future research. First, the cross-sectional design captures variable associations at a single time point, preventing the establishment of causal relationships. For example, the positive correlation between social support and the High self-compassion - high self-worth profile may stem from bidirectional influences rather than a unidirectional protective effect, which aligns with common challenges in psychological research on chronic diseases. Future longitudinal studies could employ multi-wave tracking designs to examine the dynamic evolution of these variables, clarifying temporal sequences and potential mediating pathways.

Second, regarding latent profile analysis methodology, although sensitivity analyses supported the robustness of our three-class solution, the smallest class (Class 1: Moderate self-compassion - low self-worth) comprised 18.8% of the sample. While this proportion exceeds established minimum thresholds and demonstrated stability in cross-validation, future studies with larger samples should attempt to replicate this finding to confirm the generalizability of this subgroup. Additionally, LPA results can be sensitive to indicator selection; future research could explore alternative operationalizations of self-compassion and self-worth dimensions to validate the profile structure.

Third, the sample recruitment was limited to four tertiary hospitals in Hebei Province, China, using convenience sampling, which may introduce selection bias and limit the generalizability of the results. Given that the global prevalence of TBI is higher in low- and middle-income countries, and cultural factors may moderate self-compassion expression, the geographic and cultural homogeneity of this sample may underestimate cross-cultural heterogeneity. Additionally, the higher proportion of males in the sample, predominantly urban residents, may amplify the effects of gender and urban–rural differences while neglecting the unique psychological needs of rural or minority populations.

Fourth, core variables relied on self-reported measures, which may be influenced by social desirability bias or recall bias, especially among patients with impaired cognitive function post-TBI, whose self-awareness may be limited, leading to measurement errors. Although this study ruled out severe common method bias through Harman’s single-factor test, future studies could integrate multi-method assessments, such as incorporating family reports or objective behavioral indicators, to enhance measurement validity. Fifth, potential confounding factors were not adequately controlled, such as comorbid neurological disorders, medication treatment, or quantification of trauma severity, which may mediate the association between self-compassion and self-worth. We examined death anxiety as a targeted existential construct; future studies should incorporate broader anxiety phenotypes (e.g., generalized anxiety, post-traumatic stress symptoms, fear avoidance) to determine whether different anxiety domains show distinct associations with self-compassion–self-worth profiles.

To address these limitations, future research could expand in the following directions. First, conducting longitudinal or prospective cohort studies to track the profile evolution of TBI patients from the acute to chronic phases and integrating growth mixture models to capture trajectory heterogeneity would help clarify timing and lasting effects of interventions. Second, employing multi-center, multi-cultural sample designs, such as international collaborative studies, could validate the universality and cultural specificity of profile patterns and explore the moderating effects of socioeconomic status. Third, incorporating neurobiological indicators, such as functional magnetic resonance imaging to assess prefrontal cortex activation or biomarkers (e.g., cortisol levels) as covariates, would bridge the gap between psychological heterogeneity and neurobiological mechanisms, deepening the understanding of the neurobiological basis of self-compassion. Fourth, conducting randomized controlled trials to test profile-based personalized interventions, such as comparing the efficacy of compassion-focused therapy across different subgroups and assessing cost-effectiveness to guide clinical guideline formulation. Finally, exploring advanced statistical models, such as structural equation modeling or network analysis, to examine the mediating/moderating mechanisms of death anxiety and social support, and integrating machine learning algorithms to predict profile membership risks, could lead to the development of digital screening tools, enhancing the precision of preventive interventions. These directions would not only address the shortcomings of this study but also advance the accumulation of evidence in the field of psychological rehabilitation for TBI, ultimately improving patients’ quality of life.

 Conclusion

This study revealed significant heterogeneity in the relationship between self-compassion and self-worth in patients with TBI through LPA, identifying three latent subtypes and confirming social support, death anxiety, gender, and education level as key influencing factors. These findings expand the existing literature on post-traumatic psychological heterogeneity, emphasizing the importance of examining self-perception from a person-centered perspective and providing an empirical basis for precise interventions. Specifically, by enhancing social support and alleviating death anxiety, it is possible to effectively improve patients’ levels of self-compassion, thereby buffering against the crisis of self-worth triggered by trauma and promoting overall quality of life improvement. Future research should further validate the longitudinal dynamics of these mechanisms and develop integrated models that incorporate neurocognitive indicators to promote evidence-based practices in TBI rehabilitation, meeting the needs of global patient populations.

 Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

 Ethics statement

The studies involving humans were approved by The Academic Ethics Committee of Zhangjiakou First Hospital has reviewed and approved (No.: 2025-LW-31). The studies were conducted in accordance with the local legislation and institutional requirements. The participants provided their written informed consent to participate in this study.

 Author contributions

HZ: Writing – review & editing, Data curation, Validation, Investigation, Software, Writing – original draft. LZ: Writing – review & editing, Investigation, Writing – original draft, Conceptualization, Project administration. LL: Methodology, Project administration, Investigation, Writing – review & editing, Conceptualization, Funding acquisition, Writing – original draft.

 Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

 Generative AI statement

The author(s) declared that Generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

 Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

 References GoodmanGW DevlinP WestBE RitzelRM. The emerging importance of skull-brain interactions in traumatic brain injury. Front Immunol. (2024) 15:1353513. doi: 10.3389/fimmu.2024.1353513 BertozziG MagliettaF SessaF ScotoE CipolloniL Di MizioG . Traumatic brain injury: a forensic approach: a literature review. Curr Neuropharmacol. (2020) 18:538–50. doi: 10.2174/1570159X17666191101123145, 31686630 WilsonL StewartW Dams-O'ConnorK Diaz-ArrastiaR HortonL MenonDK . The chronic and evolving neurological consequences of traumatic brain injury. Lancet Neurol. (2017) 16:813–25. doi: 10.1016/S1474-4422(17)30279-X, 28920887 HowlettJR NelsonLD SteinMB. Mental health consequences of traumatic brain injury. Biol Psychiatry. (2022) 91:413–20. doi: 10.1016/j.biopsych.2021.09.024, 34893317 MollaYD AlemuHT. Epidemiology of traumatic brain injury in Ethiopia: a systematic review and meta-analysis of prevalence, mechanisms, and outcomes. PLoS One. (2025) 20:e0322641. doi: 10.1371/journal.pone.0322641, 40446007 ZhangQ ZhangY ZhangJ. A commentary on 'management of traumatic brain injury in Africa: challenges and opportunities' (int J Surg 2024; Apr 4: online ahead of print). Int J Surg. (2024) 110:5879–80. doi: 10.1097/JS9.0000000000001639, 38752528 MahnckeHW DeGutisJ LevinH NewsomeMR BellMD GrillsC . A randomized clinical trial of plasticity-based cognitive training in mild traumatic brain injury. Brain. (2021) 144:1994–2008. doi: 10.1093/brain/awab202, 34312662 SilverbergND MikolićA. Management of psychological complications following mild traumatic brain injury. Curr Neurol Neurosci Rep. (2023) 23:49–58. doi: 10.1007/s11910-023-01251-9, 36763333 Martínez-MolinaN SiponkoskiST SärkämöT. Cognitive efficacy and neural mechanisms of music-based neurological rehabilitation for traumatic brain injury. Ann N Y Acad Sci. (2022) 1515:20–32. doi: 10.1111/nyas.14800, 35676218 DuQ LiuC LiuY LiJ GongX ZhangQ . Investigation of long-term symptoms and influencing factors in patients with mild traumatic brain injury: a cross-sectional study. Int Emerg Nurs. (2023) 69:101313. doi: 10.1016/j.ienj.2023.101313 DelmonicoRL TheodoreBR SandelME ArmstrongMA CamiciaM. Prevalence of depression and anxiety disorders following mild traumatic brain injury. PM R. (2022) 14:753–63. doi: 10.1002/pmrj.12657, 34156769 FaulknerJW SnellDL ShepherdD TheadomA. Turning away from sound: the role of fear avoidance in noise sensitivity following mild traumatic brain injury. J Psychosom Res. (2021) 151:110664. doi: 10.1016/j.jpsychores.2021.110664, 34749069 SedighimornaniN RimesKA VerplankenB. Exploring the relationships between mindfulness, self-compassion, and shame. SAGE Open. (2019) 9:2158244019866294. doi: 10.1177/2158244019866294 HalamováJ PetrovajováA ŽilinskýT. Consensual qualitative research of compassion and self-compassion definitions: a general public perspective. Qual Rep. (2022) 27:2871–84. doi: 10.46743/2160-3715/2022.5350 ColpizziI BertiC SicaC AlfeiV CaudekC. Individual differences in risk and protective factors: the role of self-compassion components among emergency responders. Behav Sci (Basel). (2024) 14:178. doi: 10.3390/bs14030178 AdonisM LoucaidesM SullmanMJM LajunenT. The protective role of self compassion in trauma recovery and its moderating impact on post traumatic symptoms and post traumatic growth. Sci Rep. (2025) 15:8145. doi: 10.1038/s41598-025-91819-x, 40059121 EganSJ ReesCS DelalandeJ GreeneD FitzallenG BrownS . A review of self-compassion as an active ingredient in the prevention and treatment of anxiety and depression in young people. Admin Pol Ment Health. (2022) 49:385–403. doi: 10.1007/s10488-021-01170-2, 34559347 MahajanSN JainA ChattopadhyayS ThemseS. Integrating rational emotive behavior therapy, compassion-focused therapy with cognitive retraining in traumatic brain injury: a case report. Consort Psychiatr. (2024) 5:93–104. doi: 10.17816/CP15546, 39980623 GirardDH KapaLL AckleyK. Comparing self-efficacy, self-compassion, and growth mindset among adults with and without traumatic brain injury. Am J Speech Lang Pathol. (2025) 34:2025–39. doi: 10.1044/2025_AJSLP-24-00327, 40315615 AshworthF GraceyF GilbertP. Compassion focused therapy after traumatic brain injury: theoretical foundations and a case illustration. Brain Impair. (2011) 12:128–39. doi: 10.1375/brim.12.2.128 CrockerJ WolfeCT. Contingencies of self-worth. Psychol Rev. (2001) 108:593–623. doi: 10.1037//0033-295X.108.3.593, 11488379 IsserowJ. Self-esteem: on the form of self-worth worth having. Pac Philos Q. (2023) 104:686–719. doi: 10.1111/papq.12434 OwnsworthT HaslamC. Impact of rehabilitation on self-concept following traumatic brain injury: an exploratory systematic review of intervention methodology and efficacy. Neuropsychol Rehabil. (2016) 26:1–35. doi: 10.1080/09602011.2014.977924, 25384260 VillaD CauserH RileyGA. Experiences that challenge self-identity following traumatic brain injury: a meta-synthesis of qualitative research. Disabil Rehabil. (2021) 43:3298–314. doi: 10.1080/09638288.2020.1743773, 32223350 MunleyPH. Erik Erikson's theory of psychosocial development and vocational behavior. J Counseling Psychol. (1975) 22:314. doi: 10.1037/h0076749 GeytenbeekM FlemingJ DoigE OwnsworthT. The occurrence of early impaired self-awareness after traumatic brain injury and its relationship with emotional distress and psychosocial functioning. Brain Inj. (2017) 31:1791–8. doi: 10.1080/02699052.2017.1346297, 29058497 DowningM ElliotB PonsfordJ. Factors influencing self-esteem after a traumatic brain injury. J Head Trauma Rehabil. (2024) 39:160–8. doi: 10.1097/HTR.0000000000000885, 37335201 KhanN RyanNP CrossleyL HearpsS CatroppaC AndersonV. Associations between peer relationships and self-esteem after childhood traumatic brain injury: exploring the mediating role of loneliness. J Neurotrauma. (2023) 40:2100–9. doi: 10.1089/neu.2022.0420, 37463051 ZhangJW ChenS Tomova ShakurTK. From me to you: self-compassion predicts acceptance of own and others' imperfections. Personal Soc Psychol Bull. (2020) 46:228–42. doi: 10.1177/0146167219853846, 31185807 ZhaoM SmithsonJ FordT WangP WongNYB KarlA. Self-compassion in Chinese young adults: specific features of the construct from a cultural perspective. Mindfulness. (2021) 12:2718–28. doi: 10.1007/s12671-021-01734-1 StephensonE WatsonPJ ChenZJ MorrisRJ. Self-compassion, self-esteem, and irrational beliefs. Curr Psychol. (2018) 37:809–15. doi: 10.1007/s12144-017-9563-2 PandeyR TiwariGK PariharP RaiPK. Positive, not negative, self-compassion mediates the relationship between self-esteem and well-being. Psychol Psychother. (2021) 94:1–15. doi: 10.1111/papt.12259, 31750614 HasanzadehF Aghaei SabetSS. The relationship between self-compassion and communication skills with social self-efficacy: the mediating role of self-worth in women with chronic pain. J Psychol Stud. (2022) 49:385–403. doi: 10.22051/PSY.2025.49795.3100 SharmaP NoronhaFS NayakAK. A correlational study to assess the relationship between body image, appearance contingent self-worth, and self-compassion among youth of Karnataka. Clin Epidemiol Glob Health. (2024) 25:101461. doi: 10.1016/j.cegh.2023.101461 BaghaipourL SajjadianI GolparvarM. Investigating the self-worth correlations of depressed women: the mediating role of acceptance, body appreciation, and psychological capital. J Tolooebehdasht. (2022), 10–18. doi: 10.18502/tbj.v21i3.10895 GatchelRJ. Psychosocial factors that can influence the self-assessment of function. J Occup Rehabil. (2004) 14:197–206. doi: 10.1023/B:JOOR.0000022761.95103.f8, 15156778 ConversanoC CiacchiniR OrrùG Di GiuseppeM GemignaniA PoliA. Mindfulness, compassion, and self-compassion among health care professionals: what's new? A systematic review. Front Psychol. (2020) 11:1683. doi: 10.3389/fpsyg.2020.01683 DreisoernerA JunkerNM Van DickR. The relationship among the components of self-compassion: a pilot study using a compassionate writing intervention to enhance self-kindness, common humanity, and mindfulness. J Happiness Stud. (2021) 22:21–47. doi: 10.1007/s10902-019-00217-4 CohenS WillsTA. Stress, social support, and the buffering hypothesis. Psychol Bull. (1985) 98:310–57. doi: 10.1037/0033-2909.98.2.310, 3901065 ThoitsPA. Social support as coping assistance. J Consult Clin Psychol. (1986) 54:416–23. doi: 10.1037//0022-006x.54.4.416, 3745593 SaanM van WeselF LeferinkS HoxJ BoeijeH van der VeldenP. Social network responses to victims of potentially traumatic events: a systematic review using qualitative evidence synthesis. PLoS One. (2022) 17:e0276476. doi: 10.1371/journal.pone.0276476, 36383517 NeffKD. Self-compassion: theory, method, research, and intervention. Annu Rev Psychol. (2023) 74:193–218. doi: 10.1146/annurev-psych-032420-031047, 35961039 LlosaJA Agulló-TomásE Menéndez-EspinaS Rivero-DíazML Iglesias-MartínezE. Self-criticism in in-work poverty: the mediating role of social support in the era of flexibility. Int J Environ Res Public Health. (2022) 19:609. doi: 10.3390/ijerph19010609, 35010871 Abd KadirN. Identity reconstruction in survivors of complex trauma. Semarak Int J Appl Psychol. (2024) 2:58–74. doi: 10.37934/sijap.2.1.5874b MenziesRE MenziesRG. Death anxiety and mental health: requiem for a dreamer. J Behav Ther Exp Psychiatry. (2023) 78:101807. doi: 10.1016/j.jbtep.2022.101807 MaasAIR MenonDK ManleyGT AbramsM ÅkerlundC AndelicN . Traumatic brain injury: progress and challenges in prevention, clinical care, and research. Lancet Neurol. (2022) 21:1004–60. doi: 10.1016/S1474-4422(22)00309-X, 36183712 RakhitS NordnessMF LombardoSR CookM SmithL PatelMB. Management and challenges of severe traumatic brain injury. Semin Respir Crit Care Med. (2021) 42:127–44. doi: 10.1055/s-0040-1716493, 32916746 RoutledgeC OstafinB JuhlJ SedikidesC CatheyC LiaoJ. Adjusting to death: the effects of mortality salience and self-esteem on psychological well-being, growth motivation, and maladaptive behavior. J Pers Soc Psychol. (2010) 99:897–916. doi: 10.1037/a0021431, 21114350 RibeiroJD HuangX FoxKR FranklinJC. Depression and hopelessness as risk factors for suicide ideation, attempts and death: meta-analysis of longitudinal studies. Br J Psychiatry. (2018) 212:279–86. doi: 10.1192/bjp.2018.27, 29587888 PyszczynskiT GreenbergJ SolomonS. A dual-process model of defense against conscious and unconscious death-related thoughts: an extension of terror management theory. Psychol Rev. (1999) 106:835–45. doi: 10.1037/0033-295x.106.4.835, 10560330 KriegerT ReberF von GlutzB UrechA MoserCT SchulzA . An internet-based compassion-focused intervention for increased self-criticism: a randomized controlled trial. Behav Ther. (2019) 50:430–45. doi: 10.1016/j.beth.2018.08.003, 30824257 EwertC VaterA Schröder-AbéM. Self-compassion and coping: a meta-analysis. Mindfulness. (2021) 12:1063–77. doi: 10.1007/s12671-020-01563-8 BeadleEJ OwnsworthT FlemingJ ShumD. The impact of traumatic brain injury on self-identity: a systematic review of the evidence for self-concept changes. J Head Trauma Rehabil. (2016) 31:E12–25. doi: 10.1097/HTR.0000000000000158, 26098262 CarrollE CoetzerR. Identity, grief and self-awareness after traumatic brain injury. Neuropsychol Rehabil. (2011) 21:289–305. doi: 10.1080/09602011.2011.555972, 21391119 YangQ ZhaoA LeeC WangX VorderstrasseA WoleverRQ. Latent profile/class analysis identifying differentiated intervention effects. Nurs Res. (2022) 71:394–403. doi: 10.1097/NNR.0000000000000597, 35417442 GabrielAS CampbellJT DjurdjevicE JohnsonRE RosenCC. Fuzzy profiles: comparing and contrasting latent profile analysis and fuzzy set qualitative comparative analysis for person-centered research. Organ Res Methods. (2018) 21:877–904. doi: 10.1177/1094428117752466 MeeusenC MeulemanB AbtsK BerghR. Comparing a variable-centered and a person-centered approach to the structure of prejudice. Soc Psychol Personal Sci. (2018) 9:645–55. doi: 10.1177/1948550617720273 YangD HuC ZhouZ HeL HuangS WanM . The impact of perceived stigma on appearance anxiety in postoperative rhinoplasty patients: a variable-centered and person-centered perspective. Acta Psychol. (2025) 260:105660. doi: 10.1016/j.actpsy.2025.105660 YangD SheT GuiG LiL ZhouZ LiuL . Psychological capital and death anxiety in pancreatic cancer patients: a latent profile analysis. Front Psych. (2025) 16:1627422. doi: 10.3389/fpsyt.2025.1627422, 41000353 SternbachGL. The Glasgow coma scale. J Emerg Med. (2000) 19:67–71. doi: 10.1016/s0736-4679(00)00182-7, 10863122 Arevalo-RodriguezI SmailagicN i FigulsMR CiapponiA Sanchez-PerezE GiannakouA . Mini-mental state examination (MMSE) for the detection of Alzheimer's disease and other dementias in people with mild cognitive impairment (MCI). Cochrane Database Syst Rev. (2015) 3:CD010783. doi: 10.1002/14651858.CD010783.pub2 NylundKL AsparouhovT MuthénBO. Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equ Model Multidiscip J. (2007) 14:535–69. doi: 10.1080/10705510701575396 RaesF PommierE NeffKD Van GuchtD. Construction and factorial validation of a short form of the self-compassion scale. Clin Psychol Psychother. (2011) 18:250–5. doi: 10.1002/cpp.702, 21584907 HuangL ChenZ JiangW QuD WangY FangX . Validation of the Chinese version of self-compassion scale for youth (SCS-Y). Mindfulness. (2022) 13:3166–78. doi: 10.1007/s12671-022-02024-0 ChenY LiuR XiaoJ WangY YangY FanH . Effects of online mindful self-compassion intervention on negative body image in breast cancer patients: a randomized controlled trail. Eur J Oncol Nurs. (2024) 72:102664. doi: 10.1016/j.ejon.2024.102664, 39059197 LubbersJ GeurtsDEM SpinhovenP Cladder-MicusMB EnnenD SpeckensAEM . Rumination and self-compassion moderate mindfulness-based cognitive therapy for patients with recurrent and persistent major depressive disorder: a controlled trial. Depress Anxiety. (2024) 2024:3511703. doi: 10.1155/da/3511703 AmbridgeJ FlemingP HenshallL. The influence of self-compassion on perceived responsibility and shame following acquired brain injury. Brain Inj. (2020) 34:945–57. doi: 10.1080/02699052.2020.1763466, 32500754 DesdentadoL CebollaA MiragallM LlorensR NavarroMD BañosRM. Exploring the role of explicit and implicit self-esteem and self-compassion in anxious and depressive symptomatology following acquired brain injury. Mindfulness. (2021) 12:899–910. doi: 10.1007/s12671-020-01553-w CrockerJ LuhtanenRK CooperML BouvretteA. Contingencies of self-worth in college students: theory and measurement. J Pers Soc Psychol. (2003) 85:894–908. ChengS-T KwanKWK. Attachment dimensions and contingencies of self-worth: the moderating role of culture. Pers Individ Differ. (2008) 45:509–14. doi: 10.1016/j.paid.2008.06.003 GengL JiangT. Contingencies of self-worth moderate the effect of specific self-esteem on self-liking or self-competence. Soc Behav Personal Int J. (2013) 41:95–107. doi: 10.2224/sbp.2013.41.1.95 ValikhaniA GoodarziMA. Contingencies of self-worth and psychological distress in Iranian patients seeking cosmetic surgery: integrative self-knowledge as mediator. Aesth Plast Surg. (2017) 41:955–63. doi: 10.1007/s00266-017-0853-8, 28374299 García-SorianoG BellochA. Exploring the role of obsessive-compulsive relevant self-worth contingencies in obsessive-compulsive disorder patients. Psychiatry Res. (2012) 198:94–9. doi: 10.1016/j.psychres.2011.11.011, 22386566 GedikZ IdimanE. Health-related quality of life in multiple sclerosis: links to mental health, self-esteem, and self-compassion. J Psychiatry Neurol Sci. (2020) 33. doi: 10.14744/DAJPNS.2019.00061 ZimetGD DahlemNW ZimetSG FarleyGK. The multidimensional scale of perceived social support. J Pers Assess. (1988) 52:30–41. doi: 10.1207/s15327752jpa5201_2 YangX XueM PauenS HeH. Psychometric properties of the Chinese version of multidimensional scale of perceived social support. Psychol Res Behav Manag. (2024) 17:2233–41. doi: 10.2147/PRBM.S463245, 38835653 ZengEQ ZengBQ TianJL DuB TianXB ChenH. Perceived social support and its impact on mental fatigue in patients with mild traumatic brain injury. Balkan Med J. (2016) 33:152–7. doi: 10.5152/balkanmedj.2016.15701, 27403383 TramontiF BonfiglioL Di BernardoC UliviC VirgillitoA RossiB . Family functioning in severe brain injuries: correlations with caregivers' burden, perceived social support and quality of life. Psychol Health Med. (2015) 20:933–9. doi: 10.1080/13548506.2015.1009380, 25668723 TemplerDI. The construction and validation of a death anxiety scale. J Gen Psychol. (1970) 82:165–77. doi: 10.1080/00221309.1970.9920634 YangH ZhangJ LuY LiM. A Chinese version of a Likert-type death anxiety scale for colorectal cancer patients. Int J Nurs Sci. (2016) 3:337–41. doi: 10.1016/j.ijnss.2016.11.002 YanY ChenY OuM GongY YangR LiuX . The mediating role of meaning in life between experiential avoidance and death anxiety among cancer patients: a cross-sectional study. BMC Cancer. (2024) 24:663. doi: 10.1186/s12885-024-12433-0, 38822257 Quevedo-BlascoR Díaz-RománA Vega-GarcíaA. Death anxiety in caregivers of chronic patients. Healthcare (Basel). (2024) 12:107. doi: 10.3390/healthcare12010107 FirooziM RouhiS. Does fear impede the success of spinal surgery in patients with chronic back pain? Potential prognosticative factors in spinal surgery failure. Curr Psychol. (2023) 42:7414–20. doi: 10.1007/s12144-021-02069-9 KlineRB. Principles and practice of structural equation modeling. New York, NY: Guilford Publications (2023). Nylund-GibsonK ChoiAY. Ten frequently asked questions about latent class analysis. Transl Issues Psychol Sci. (2018) 4:440–61. doi: 10.1037/tps0000176 LubkeGH MuthénB. Investigating population heterogeneity with factor mixture models. Psychol Methods. (2005) 10:21–39. doi: 10.1037/1082-989X.10.1.21, 15810867 WeiL XieJ WuL YaoJ ZhuL LiuA. Profiles of self-compassion and psychological outcomes in cancer patients. Psychooncology. (2023) 32:25–33. doi: 10.1002/pon.5931, 35334138 AkiM ShibataM FujitaY SpantiosM KobayashiK UenoT . Neural basis of self-esteem: social cognitive and emotional regulation insights. Front Neurosci. (2025) 19:1588567. doi: 10.3389/fnins.2025.1588567, 40463590 ZhangH CarrER Garcia-WilliamsAG SiegelmanAE BerkeD Niles-CarnesLV . Shame and depressive symptoms: self-compassion and contingent self-worth as mediators? J Clin Psychol Med Settings. (2018) 25:408–19. doi: 10.1007/s10880-018-9548-9, 29488038 da Silva-SauerL de la Torre-LuqueA Basso GarciaR Fernández-CalvoB. Self-kindness as a protective factor: mitigating the impact of social restriction on mental health of family caregivers during the COVID-19 pandemic. Aging Ment Health. (2025) 29:1356–65. doi: 10.1080/13607863.2025.2489711, 40220339 ZhouL SukpasjaroenK WuY GaoL ChankosonT CaiE. Perceived social support promotes nursing students' psychological wellbeing: explained with self-compassion and professional self-concept. Front Psychol. (2022) 13:835134. doi: 10.3389/fpsyg.2022.835134, 35478770 TombergT ToomelaA PulverA TikkA. Coping strategies, social support, life orientation and health-related quality of life following traumatic brain injury. Brain Inj. (2005) 19:1181–90. doi: 10.1080/02699050500150153, 16286333 LairdV ElliottTR BrossartDF LuoW HicksJA WarrenAM . Trajectories of affective balance 1 year after traumatic injury: associations with resilience, social support, and mild traumatic brain injury. J Happiness Stud. (2019) 20:1419–37. doi: 10.1007/s10902-018-0004-1 FairlambS JuhlJ. Death thoughts predict increased depression for those with low self-worth. Death Stud. (2022) 46:1021–6. doi: 10.1080/07481187.2020.1793432, 32673184 YarnellLM NeffKD DavidsonOA MullarkeyM. Gender differences in self-compassion: examining the role of gender role orientation. Mindfulness. (2019) 10:1136–52. doi: 10.1007/s12671-018-1066-1 AlbertsonER NeffKD Dill-ShacklefordKE. Self-compassion and body dissatisfaction in women: a randomized controlled trial of a brief meditation intervention. Mindfulness. (2015) 6:444–54. doi: 10.1007/s12671-014-0277-3 MusabiqS SantosoGPA TakwinB PoerwandariEK. Differential pattern of consequences of self-compassion across gender: predicting flourishing and life satisfaction by two factors of self-compassion. J Theor Soc Psychol. (2024) 2024:5564391. doi: 10.1155/2024/5564391 GaoY ZhengY HeY LinJ LiuF FuJ . The relationship between self-efficacy, health literacy, and quality of life in patients with chronic diseases: a cross-sectional study in China. Front Public Health. (2024) 12:1430202. doi: 10.3389/fpubh.2024.1430202, 39391157 SelzlerAM HabashR RobsonL LentonE GoldsteinR BrooksD. Self-efficacy and health-related quality of life in chronic obstructive pulmonary disease: a meta-analysis. Patient Educ Couns. (2020) 103:682–92. doi: 10.1016/j.pec.2019.12.003, 31859120 KitayamaS MarkusHR. Culture and the self: implications for cognition, emotion, and motivation. Psychol Rev. (1991) 98:224–53. doi: 10.1037/0033-295x.98.2.224 LuL GilmourR. Culture and conceptions of happiness: individual oriented and social oriented SWB. J Happiness Stud. (2004) 5:269–91. doi: 10.1007/s10902-004-8789-5 WongYJ. Strength-centered therapy: a social constructionist, virtues-based psychotherapy. Psychother Theory Res Pract Train. (2006) 43:133. doi: 10.1037/0033-3204.43.2.133, 22122032 YanY. The individualization of Chinese society. London, UK: Routledge (2020).

Edited by: Yari Longobucco, University of Florence, Italy

Reviewed by: Monica Cornea, Dr. Gheorghe Preda Clinical Psychiatric Hospital of Sibiu, Romania

Yingying Gu, Nanchang University, China