Depression, anxiety and stress in adolescents and their associated factors: a latent profile analysis

Abstract

Background

Most research on mental health in adolescent centers on anxiety, depression, and their predictors. The characteristics of depression, anxiety, and stress symptoms in adolescents from an overall viewpoint remains unclear.

Objective

To understand the characteristics of depression, anxiety, and stress in different subgroups of adolescents and analyze the influencing factors.

Method

A cross-sectional study was conducted using the Depression Anxiety Stress Scales for Youth. Potential characteristics and related influencing factors were analyzed using latent profile analysis, a regression mixed model, and multiple logistic regression.

Results

A total of 1408 higher vocational college students aged 14 ~ 17 years were included in this study. A two-profile model was proposed as the best: a High DASS-Y group (8.95%) with moderate and above negative emotions and a Low DASS-Y group (91.05%) with normal levels of negative emotional symptoms. Negative life events in adolescents could be significantly predicted, supporting the model’s reliability. Adolescents from multi-child families (OR = 1.664, 95% CI: 1.077 ~ 2.571), those with chronic diseases (OR = 11.505, 95% CI: 6.354 ~ 20.840) and those with academic performance below 50% were at higher risk of negative emotions (OR = 1.705, 95% CI: 1.024 ~ 2.837). Adolescents who majored in science had a lower risk of negative emotions (OR = 0.513, 95% CI: 0.307 ~ 0.857).

Conclusions

These findings suggest that most of the negative emotional symptoms of adolescents were within the normal range. Although the High DASS-Y group was in a minority, the symptoms of depression, anxiety and stress were more serious, and more attention should be paid to adolescents, especially those from multi-child families, with chronic diseases and poor academic performance. Schools and parents should further provide relevant mental health education and strengthen emotional support for adolescents.

Clinical trial number

Not applicable.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12888-025-07719-x.

Introduction

The World Health Organization estimates that 50% of mental health problems occur in children and adolescents [1], which raises the risk of self-harm, drinking, smoking, and violent conduct. It also accounts for the majority of youth disease burden [2]. In China, 14.8% and 15.8% of adolescents have different degrees of risk of depression and anxiety, respectively, which are higher than those of adults [3]. Moreover, anxiety and depression are consistently identified as the most common mental health concerns among middle and high school students [4], who also report higher stress levels and are more likely to engage in non-suicidal self-injury than adults [5]. Despite the high prevalence, fewer than half of affected adolescents receive timely treatment, resulting in delayed academic progression, increased family burden, and substantial societal costs [6, 7].

Adolescents are experiencing rapid physical and psychological development, with exceptional personalities, strong self-awareness, a desire for attention, and curiosity. However, they are still immature and susceptible, making this the peak period for the occurrence and development of negative emotions, such as depression, anxiety, and stress [8]. Several studies have shown that negative emotions in adolescents are often associated with negative experiences, including poor academic performance, social problems, family conflict, physical illness, and other negative life events [5, 7, 9]. Moreover, empirical studies have shown that exposure to negative life events can increase the risk of developing mental health problems among adolescent [10, 11]. These effects might persist for several years [12]. Therefore, it is crucial to identify not only the population characteristics of adolescent depression, anxiety, and stress, but also the influence of negative life events, so as to design tailored early interventions and prevent the progression and severity of mental health disorders.

Currently, the majority of research on mental health issues in adolescents focuses on anxiety or depression symptoms and the factors that predict them [13, 14]. Few studies have investigated the characteristics of depression, anxiety, and stress symptoms in adolescents from an overall viewpoint, even though these symptoms frequently coexist and potentially predict one another [15–18]. It will be easier to avoid and intervene in the onset and development of negative emotions in high-risk adolescents if the features and contributing causes of depression, anxiety, and stress symptoms in adolescents are recognized. Children and adolescents between the ages of 8 and 17 can benefit from the Depression Anxiety Stress Scales for Youth (DASS-Y) [19], which is an excellent tool for grading the severity of the main symptoms of depression, anxiety, and stress. It has strong psychometric qualities and aids in the study’s overall evaluation of negative emotions in adolescents.

Latent profile analysis (LPA) is a statistical method that classifies individuals according to their different response patterns on the observed indicators to identify population heterogeneity [20]. The models used in latent class analysis (LCA) and LPA are collectively referred to as latent class models. LCA is applied to categorical indicators and identifies groups based on distinct response probabilities, whereas LPA is used for continuous indicators, identifying profiles that differ in their mean levels across variables. Because depression, anxiety, and stress are typically measured as continuous symptom dimensions in the DASS-Y, using LPA provides a more precise representation of individual differences in symptom intensity, whereas discretizing continuous variables for LCA would inevitably lead to information loss and reduced statistical efficiency [21]. A systematic review and meta-analysis showed [22] that generalizable interventions were less effective than targeted mental health programs in reducing adolescent depression, anxiety, and stress. With the use of LPA, we can better understand differences in adolescents’ emotional symptom patterns by examining how sociodemographic factors relate to depression, anxiety, and stress, and thus to design more targeted interventions. Therefore, this study will comprehensively evaluate adolescents’ depression, anxiety, and stress, and explore the characteristics of adolescent negative emotions through LPA to determine the influencing factors of their optimal potential profiles. To further validate the LPA classification, characterize the subgroups of adolescents with depression, anxiety, and stress, and provide evidence for the construct validity of these profiles, this study investigated negative life events between LPA subgroups.

In summary, this study employed LPA to identify different aspects of adolescents’ depression, anxiety, and stress. The secondary goal was to examine the predictive power of sociodemographic features within the negative emotions subgroup. Lastly, the validity of the latent profile model was confirmed using adverse occurrences in adolescents’ lives.

Methods

Design, setting and procedure

The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist was followed in the reporting of this cross-sectional study [23]. This cross-sectional study was conducted between June and September 2023. We employed convenience sampling to recruit first and second-year students from two higher vocational colleges in Suzhou, located in eastern China. The study was approved by the Soochow University Ethics Committee (No. SUDA20230428H15). After the two schools agreed to recruit the study participants, the researchers determined the unified guidance language, informed the participants about the purpose of the study, how to fill in the survey, and its anonymity, and obtained informed consent. The questionnaires were returned on the spot immediately.

Participants

Inclusion criteria for this study were as follows: (a) 8 to 17 years of age; (b) no language or communication impairment and good comprehension. Exclusion criteria included students who were unwilling to cooperate. A total of 1500 students were surveyed, and 1408 students were finally included after excluding cases with > 5% missing data (valid response rate = 93.87%).

Measurements

Basic information

Based on literature review and expert consultation, a self-designed general information questionnaire was developed. The sociodemographic data mainly included age, sex, ethnicity, major, place of origin, academic performance, family per capita income, single-parent family, one-child family and health status.

Depression Anxiety Stress Scales for Youth (DASS-Y) [19]: We used the Chinese version of DASS-Y, which had been sinicized and validated by our research team in June 2023 [24]. The final version consists of three subscales with 21 items that assess anxiety (seven items), depression (seven items), and stress (seven items) in the past week. A Likert 4-point scale (0–3 points) was used, with higher scores indicating more serious negative emotions, and negative emotions could be comprehensively evaluated. Total and subscale scores can be classified as normal, mild, moderate, severe, and extremely severe, and the specific cut-off values can be found on the DASS website [25]. The Cronbach’s α coefficient for this questionnaire was 0.961.

Positive affect and negative affect scale for Children (PANAS-C) [26]: The revised Chinese version of PANAS-C scale was used to measure the positive and negative affect of middle school students. The positive affect subscale (PA) and the negative affect subscale (NA) comprised 27 items. The scale was a 5-point scale (1–5 points), and the average score of items was used as the emotional evaluation index. The higher the score, the stronger the corresponding emotion. In this study, the Cronbach’s α value of this questionnaire was 0.943.

Adolescent Self-Rating Life Events Check-list (ASLEC) [27]: It is composed of 27 negative life events that may cause psychological and physiological reactions to adolescents. The response to each event was to determine whether it had occurred during the past week. If it had, the psychological feelings at the time of the event were graded into 5 levels: no impact (1), mild (2), moderate (3), severe (4), and extremely severe (5). Higher scores indicated a more substantial impact of life events on the participants, and the observation index was calculated as the sum of the total values for each dimension. The Cronbach’s α coefficient in this study was 0.971.

Statistical analyses

Excel 2016 was used to establish a database for data entry and verification by two people. Questionnaires with missing values > 5% were excluded, and other missing values were filled with the median [28]. SPSS 26.0 was used for descriptive statistical analysis and multiple logistic regression. Mplus 8.3 was used for latent profile analysis, and the latent profile models of depression, anxiety, and stress levels were constructed based on the scores of the three dimensions of DASS-Y. The model testing indexes included Akaike information criterion (AIC), Bayesian information criterion (BIC) and Adjusted BIC (ABIC), Entropy, Lo-Mendell-Rubin (LMR) and Bootstrapped Likelihood Ratio Test (BLRT). The smaller the AIC, BIC, and ABIC, the better the fitting degree of the model [29]. The closer the entropy value is to 1, the better the fitting degree of the model is [30]. LMR and BLRT can compare the fitting effect of two adjacent models, and P < 0.05 indicates that the k model has a better fitting effect than the k-1 model, with k equaling the number of classes [31]. In addition, the proportion of people in each category should be more than 5%, otherwise the classification is not reasonable [32]. Cohen’s d illustrates the standardized magnitude differences of the dimensions of DASS-Y in the LPA profile, and values of d < 0.20 indicate negligible differences, 0.20–0.50 small effects, 0.50–0.80 moderate effects, and d ≥ 0.80 represent large effects [33]. To test the influence of sociodemographic data on the latent profile of adolescents’ depression, anxiety, and stress levels and control the classification error, a three-step method was used based on a regression mixture model for subsequent analysis with Mplus. The R3STEP command in the AUXILIARY option [20] was used, with the LPA analysis results as the dependent variable and the sociodemographic data as the predictor, to explore its impact on the negative emotions of junior college students. To confirm the profiles of adolescent depression, anxiety, and stress, multiple logistic regression was finally carried out in SPSS 26.0. This involved examining the predictive power of ASLEC scores using DASS-Y categories as dependent factors and PA, NA, and ASLEC as independent variables.

Results

Participant characteristics

A total of 1408 higher vocational college students aged 14 ~ 17 years were included in this study, mainly male (72.59%). Complete participants information is provided in Table 1.

Table 1.

Participants’ sociodemographic characteristics (N = 1408)

Characteristics	N (%)	Characteristics	N (%)
Sex		Place of origin
Male	1022 (72.59)	Rural area	658 (46.73)
Female	386 (27.41)	City area	750 (53.27)
Age (year)		Academic performance
14 ~ 15	214 (15.20)	Bottom 50%	334 (23.72)
16	996 (70.74)	30% ~50%	348 (24.72)
17	198 (14.06)	15% ~30%	260 (18.47)
Ethnic minorities		Top 15%	466 (33.09)
Yes	34 (2.41)	Physical condition
No	1374 (97.59)	Healthy	1356 (96.31)
Major		Have a chronic illness	52 (3.69)
Literature and art	194 (13.78)	Single parent family
Science	428 (30.40)	Yes	110 (7.81)
Engineering	786 (55.82)	No	1298 (92.19)
Monthly per capita family income		One-child Family
< 5000 yuan	416 (29.55)	Yes	522 (37.07)
5000 ~ 7000 yuan	460 (32.67)	No	886 (62.93)
> 7000 yuan	532 (37.78)

Latent profiles determination

Table 2 lists the fit indices of latent profile models for classes 1 to 6. Although the fit indices continued to decline with increasing profile numbers, the two-profile model demonstrated the most substantial improvement in AIC, BIC, and ABIC values (AIC = 22484.77, BIC = 22537.27, ABIC = 22505.50). In addition, the two-profile solution showed the highest entropy value (0.983), indicating excellent classification accuracy. Both the LMR (P <0.001) and BLRT (P <0.001) tests were significant, supporting that the two-profile model provided a significantly better fit than the one-profile model. The category proportions of model 2 were 91.05% and 8.95%, both of which were greater than 5%. Therefore, the two-profile model is the optimal model.

Table 2.

Model fitting indexes for LPA in DASS-Y

Profiles	m	AIC	BIC	ABIC	Entropy	LMR(P)	BLRT(P)	Profile prevalence (%)
One-profile	6	25012.59	25044.09	25025.03
Two-profile a	10	22484.77	22537.27	22505.50	0.983	< 0.001	< 0.001	8.95/91.05
Three-profile	14	21242.91	21316.41	21271.93	0.939	0.0065	< 0.001	21.73/72.73/5.54
Four-profile	18	20575.95	20670.44	20613.27	0.952	< 0.001	< 0.001	22.16/ 69.46/5.82/2.56
Five-profile	22	20031.13	20146.63	20076.74	0.966	< 0.001	< 0.001	69.60/15.77/6.96/2.41/ 5.26
Six-profile	26	19830.48	19966.98	19884.39	0.956	0.0867	< 0.001	66.90/2.98/8.24/4.97/14.35/2.56

Note. m: the number of free parameters; AIC: Akaike Information Criterion; BIC: Bayesian Information Criterion; ABIC: Adjusted Bayesian Information Criterion; LMR: Lo-Mendell-Rubin; BLRP: Bootstrapped Likelihood Ratio Test; DASS-Y: Depression Anxiety Stress Scales For Youth

^a Optimal model

The probabilities of class 1 and class 2 in the two-profile model were 97.9% and 99.8%, respectively, indicating good discrimination and reliability (see Table A1). Combining Fig. 1; Table 3, it was observed that class 1 scored significantly higher than class 2 in all three dimensions of DASS-Y (d_2 − 1 = 2.86 ~ 3.29), which indicates large effect sizes. According to the naming convention, class 1 was named High DASS-Y group and class 2 was named Low DASS-Y group.

Fig. 1.

The mean scores of the three dimensions of DASS-Y in the two profiles. Note. DASS-Y: Depression Anxiety Stress Scales For Youth

Table 3.

Mean, standard Deviation, and d for DASS-Y subscales of the two profiles of DASS-Y

	Class-1 (N = 126)			Class-2 (N = 1282)
Variables	Mean	SD	Range of scores	Mean	SD	Range of scores	d2 − 1
Depression	14.56	4.52	3 ~ 21	1.93	3.01	0 ~ 18	3.29
Anxiety	13.11	5.26	3 ~ 21	1.48	2.30	0 ~ 12	2.86
Stress	15.44	4.06	8 ~ 21	3.51	3.55	0 ~ 19	3.13

Note. The mean scores of three subscales of DASS-Y are provided. DASS-Y: Depression Anxiety Stress Scales For Youth; Class-1 = High DASS-Y group; Class-2 = Low DASS-Y group

Influencing factors of latent categories

Prior to the logistic regression analyses, multicollinearity was assessed using VIF, and no concerns were found (all VIFs <2). Table 4 shows the results of the robust three-step method and whether being an only child, health status, major, and academic performance have significant predictive effects on the latent profile of DASS-Y. Specifically, adolescents from multi-child families had a higher risk of depression, anxiety, and stress symptoms compared with the Low DASS-Y group (OR = 1.664, 95% CI: 1.077 ~ 2.571). Adolescents with chronic medical conditions were at higher risk for symptoms of depression, anxiety, and stress (OR = 11.505, 95% CI: 6.354 ~ 20.840). The risk of depression, anxiety, and stress symptoms was higher in adolescents with academic performance ranking below 50% (OR = 1.705, 95% CI: 1.024 ~ 2.837). Science-major adolescents had a lower risk of depression, anxiety, and stress symptoms (OR = 0.513, 95% CI: 0.307 ~ 0.857).

Table 4.

Results of multivariate logistic regression analysis (N = 1408)

Variable	High DASS-Y group a
		95%CI
	OR	Low	Up	P
Sex (ref. Male)
Female	0.575	0.280	1.181	0.132
Age	0.758	0.502	1.144	0.187
Place of origin (ref. Rural area)
City area	1.281	0.852	1.926	0.233
Single parent family (ref. Yes)
No	0.568	0.301	1.073	0.081
One-child Family (ref. Yes)
No	1.664	1.077	2.571	0.022
Physical condition (ref. Healthy)
Have a chronic illness	11.505	6.354	20.840	< 0.001
Monthly per capita family income (ref. >7000 yuan)
< 5000 yuan	0.828	0.498	1.375	0.466
5000 ~ 7000 yuan	0.802	0.503	1.278	0.354
Academic performance (ref. > Top 15%)
Bottom 50%	1.705	1.024	2.837	0.040
30% ~50%	0.982	0.568	1.697	0.948
15% ~30%	0.669	0.325	1.376	0.275
Major (ref. Engineering)
Literature and art	0.939	0.438	2.013	0.871
Science	0.513	0.307	0.857	0.011

Note. ^a Reference Low DASS-Y group; DASS-Y: Depression Anxiety Stress Scales For Youth

Examining the best-fitting profile using PA, NA, and ASLEC predictors

Multinomial logistic regression was performed using the DASS-Y latent profile as the outcome variable, the Low DASS-Y group as the reference group, and PA, NA, and ASLEC as predictors (Table 5). The results showed that for each point increase in ASLEC in this group, the likelihood of falling in the High DASS-Y group increased by 2.4% compared to the Low DASS-Y group. In the same way, the probability of falling into the High DASS-Y group increased by 8.766 times for every point increase in NA. However, the probability of falling into the High DASS-Y group decreased as PA increased.

Table 5.

Multinomial logistical regression results of PA, NA and ASLEC on the two-profile model

Variables	ASLEC		PA		NA
	Class-1	Class-2	Class-1	Class-2	Class-1	Class-2
Mean ± SD	55.08 ± 43.51	15.25 ± 19.86	2.98 ± 1.30	2.96 ± 0.98	3.22 ± 0.98	1.63 ± 0.72
OR	1.024		0.338		9.766
95% CI	(1.016–1.032)		(0.226–0.505)		(6.233–15.302)
P	< 0.001		< 0.001		< 0.001

Note. Low DASS-Y group is the reference group. ASLEC, Adolescent Self-Rating Life Events Check-list; PA, Positive affect scale; NA, Negative affect scale. Class-1 = High DASS-Y group; Class-2 = Low DASS-Y group

Discussion

By employing latent profile analysis, we identified two distinct profiles of adolescents’ negative emotions and examined their potential sociodemographic risk factors. To our knowledge, this is the first application of the DASS-Y among Chinese adolescents. This provides preliminary evidence supporting the cultural applicability of the DASS-Y in the Chinese context and contributes to understanding group differences in adolescents’ negative emotions from an overall viewpoint encompassing depression, anxiety, and stress, as well as their potential implications for intervention.

This study surveyed 1,408 Chinese adolescents and reported prevalence of negative emotions, anxiety and stress of 76.86%, 50.28%, and 73.72%, respectively—similar to findings reported by Liu et al. [34], and aligned with developmental psychopathology theory [35]. Furthermore, latent profile analysis identified two distinct emotional profiles: the Low DASS-Y group and the High DASS-Y group. According to the severity cut-off value in the DASS-Y manual [25], although only a minority of adolescents fell into the High DASS-Y group, their anxiety and depression scores were classified as severe, with stress symptom scores in the moderate range. In contrast, all scores for the Low DASS-Y group remained within the normal range. This indicates that the overall mental health status of adolescents is relatively stable, but the identification and support of high-risk groups are particularly crucial. Therefore, schools and families should implement routine mental health monitoring systems to facilitate proactive prevention and early identification. Schools can offer mental health courses to provide preventive promotion for low-risk students. High-risk students should receive focused attention with timely intervention, ensuring professional medical involvement when necessary.

According to the stress-vulnerability model, the interaction between multiple environmental stressors and existing personal vulnerabilities may increase adolescents’ risk of belonging to the High DASS-Y group [36]. Our findings indicate that multi-child families, chronic illness, and academic performance ranking below 50% are primary risk factors, while positive emotions exert a protective effect. However, some factors previously identified as significant predictors of negative emotions in adolescents were not statistically significant in this study. For instance, prior research has shown that females are more prone to negative emotions [37] and that adolescents with divorced parents experience higher levels of depression, anxiety, and stress [38]. The inconsistencies may be attributed to the uneven distribution of these variables, which may have influenced the statistical results, as only 27.41% of participants were female and 7.81% were from single-parent families. The use of LPA for population classification and the adoption of a two-profile model as the optimal solution may also be one of the factors. Furthermore, they may also reflect differences in sample characteristics, regional context, or data collection timing, as well as potential self-reporting bias.

Chronic illness, as a stressor factor, impacts adolescents’ mental health. In this study, adolescents with chronic illnesses were more likely to belong to the High DASS-Y group than their healthy peers. However, only 3.69% of participants reported a chronic illness in this study, which may have affected the stability of the results. Even so, this finding aligns with prior research [39–41] indicating strong associations between chronic physical illnesses and psychological distress. Potential explanations include disease-related physical symptoms (e.g., fatigue, pain), medication side effects, and activity limitations that compromise quality of life and social participation. Additionally, adolescents may engage in pathological rumination about disease prognosis and mortality, triggering intense negative emotions [42, 43]. Implementing cognitive behavioral therapy can improve adolescents’ emotional states by altering thoughts and behaviors. Furthermore, flexible community support programs should be promoted, integrating mental health services with disease knowledge education and family support to enhance psychological resilience while managing illness [44].

Bronfenbrenner’s Ecological System Theory proposes that the family and school are crucial microsystems influencing adolescent development, with the family atmosphere and school resources closely associated with psychological growth [45]. In China, the traditional emphasis on academic performance represents an external pressure that significantly increases adolescents’ psychological burden. Particularly, high school students also face pressures from college entrance exams, career and major choices, and parental expectations [46]. Excessive academic pressure has been proven to be closely associated with adolescent anxiety and sleep disorders [47]. This study’s findings indicate that adolescents in the bottom 50% of academic rankings face higher risks of depression, anxiety, and stress symptoms, while science students exhibit relatively lower risks—consistent with Jiang et al. [37]. This may reflect the sense of accomplishment from subject strengths, whereas underperforming adolescents are more prone to helplessness and self-denial. Family understanding and support can partially buffer the negative effects of academic pressure [48]. Parents are advised to avoid focusing solely on grade improvement and instead establish reasonable expectations as well as prioritize their children’s mental health by adopting scientific educational approaches and stress-reduction strategies. Additionally, schools should establish diversified evaluation systems while maintaining academic standards, reducing reliance on grade rankings as the sole core criterion to minimize peer comparison and diminished self-efficacy. At the mesosystem level, strengthening collaboration between families and schools is equally important. Establishing regular mechanisms for mental health monitoring and communication, such as school-family meetings, can facilitate information sharing and coordinated intervention efforts.

Additionally, the family environment plays a fundamental role within the microsystem. This study found that multi-child families constitute a risk factor for adolescents belonging to the High DASS-Y group. This stems from the division of parental attention and the complexity of family resource allocation, resulting in reduced emotional support for each child. Consequently, children may compete more intensely for parental recognition. Insufficient support conflicts with adolescents’ developmental needs for autonomy and identity formation, exacerbating their psychological vulnerability [37]. Therefore, family education should emphasize procedural fairness and effective communication, provide timely emotional support tailored to children’s needs, and foster sibling cooperation through collaborative family tasks and group rewards.

According to the stress-vulnerability model, repeated negative life experiences may heighten adolescents’ vulnerability to emotional disorders, particularly around puberty [10]. This study validated the predictive role of negative life events on depression, anxiety, and stress through multiple logistic regression, consistent with prior research [11, 48]. As Zheng et al. [39] noted, negative life events serve as significant environmental predictors of suicidal ideation and behavior. The cognitive appraisal theory of stress suggests that when individuals feel they cannot handle stress and expect negative outcomes, they are likely to show negative emotional reactions such as anxiety, avoidance, and tension [49]. Influenced by traditional Chinese cultural values like “self-restraint” and “introspection”, Chinese adolescents often choose to hide negative emotions rather than express them or ask for help [48]. They also tend to make internal attributions during stressful situations, which can lead to excessive rumination. These patterns make it harder to correctly evaluate and cope with stress, and may increase the risk of depression, anxiety, and lower well-being. These findings suggest that parents and educators should promptly attend to adolescents’ stress experiences, foster supportive environments, and assist them in developing effective stress coping and emotion regulation strategies. Meanwhile, leveraging positive emotions as a protective factor, such as through positive psychology training [50], can mitigate the risks associated with adverse life events.

Limitations

First of all, although this study identified associations between negative emotions and multiple risk factors using a cross-sectional design, causal relationships cannot be inferred. Future research should adopt longitudinal or prospective cohort designs to clarify temporal and causal pathways. Second, convenience sampling was used and the sample was limited to students from two schools in Suzhou. This may have introduced selection bias and reduced representativeness. Unmeasured cultural factors, such as stigma and social desirability [51, 52], may lead adolescents to under-report psychological symptoms, thereby biasing the findings toward underestimation. Multi-center designs with larger and more diverse samples are encouraged in further research. In addition, the sample consisted only of higher vocational college students aged 14–17, and future research should include younger age groups across different educational settings (e.g., junior high and primary school) to examine the generalizability of the findings in this study. Moreover, all data were self-reported, which may have introduced response and recall bias. Therefore, future research should incorporate objective measures to identify the causal mechanism of negative emotions in adolescents.

Conclusion

This study used latent profile analysis to identify the High DASS-Y group and the Low DASS-Y group based on the three dimensions of DASS-Y. There were significant differences in negative emotion scores between the two groups. The Low DASS-Y group contained most adolescents whose negative emotions were within the normal range. While the High DASS-Y group accounted for a minority, its negative emotions were severe, and it was the group that society, schools, and families focused on, especially the students from multi-child families, and those with chronic diseases or poor academic performance. The findings may help mental health practitioners, schools, and parents to identify adolescents with high levels of negative emotional symptoms at an early stage and classify them for more targeted prevention and intervention efforts.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the study’s conception and design. Material preparation, data collection and analysis were performed by Z.L., L.J.L., Q.C., and J.W. The first draft of the manuscript was written by Z.L., and L.J.L., and was supervised by Z.F.Y., and L.T. All authors read and approved the final manuscript.

Funding

This study was supported by the Major Project of Philosophy and Social Science Research in Colleges and Universities in Jiangsu Province (No. 2023SJZD144). They had no role in the study design, collection, analysis, or interpretation of the data, writing the manuscript, or the decision to submit the paper for publication.

Data availability

All data generated or analyzed during the current study are included in this article.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Soochow University (Approval No. SUDA20230428H15). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. Informed consent was obtained from all individual participants included in the study.

Consent for publication

Not applicable.

Submission declaration

This manuscript has not been published and is not under consideration for publication elsewhere. All authors have read and approved the final version of the manuscript.

Competing interests

The authors declare no competing interests.