Relationships between depression and difficulties in emotion regulation among first-year college students: a network analysis approach

Abstract

Background

Depression and difficulties in emotion regulation (DER) may co-occur in first-year college students due to the transition from high school to college environment. However, the intricate interaction dynamics between depression and difficulties in emotion regulation symptoms are unclear. This study employed network analysis to examine the network structure of depression and difficulties in emotion regulation among first-year college students.

Methods

This cross-sectional study included nine hundred and ninety-two first-year Chinese college students (M_age = 18.68, SD = 0.85) who completed the Patient Health Questionnaire and Difficulties in Emotion Regulation Scale.

Results

“Lack of emotional clarity” and “non-acceptance of emotional responses” emerged as bridge symptoms for the network. The strongest connections are between “non-acceptance of emotional reactions” and “limited access to effective emotion regulation strategies”, “impulse control difficulties”, and “difficulties engaging in goal-directed behavior”, respectively. Network structure and global strength did not differ by gender, but some edge weights varied.

Conclusion

These findings can inform the development of interventions targeting comorbid depression-DER onset among transitioning college students.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40359-025-03198-7.

Introduction

The first year of college is a significant transition point from the high school to the college environment [1]. This period poses significant challenges for first-year students as they may encounter various stressors including academic pressures [2], complex interpersonal relationships [3], and independent living [4]. These transitions can lead to psychological problems such as depression [5], with previous studies revealing that 45.3% of first-year Chinese college students reported depressive symptoms, and 6.33%-6.60% experiencing moderate to severe symptoms [6, 7]. Depression not only affects students’ mental health but also increases the risk of obesity [8], academic difficulties [9], substance abuse [10], and suicidal ideation [11]. Hence, it is pivotal to investigate the risk factors underlying the development of depressive symptoms. One of the key factors identified in research is the difficulty in emotion regulation (DER), which has been shown to be a significant predictor of depression and other mental health issues [12].

Difficulties in emotion regulation (DER) is noticeable in several clinical diseases as well as a wide range of emotional problems [12–14]. Many studies have demonstrated that deficiencies in the regulation of emotions are widely recognized as a significant risk factor for the development of depression [15–17]. Gratz and Romer [18] argue that emotion regulation involves a variety of components, including (a) emotional awareness and comprehension, (b) emotional acceptance, (c) ability to control impulsive behaviors and behave in accordance with desired goals when experiencing negative emotions, and (d) ability to use situationally appropriate emotion regulation strategies. A lack of proficiency in any of these skills indicates emotional regulation difficulties. According to this conceptualization, Bjureberg et al. [19] developed a brief version of the Difficulties in Emotion Regulation Scale (DERS-16) to enhance the scale’s clinical and research applicability. It covers five dimensions: lack of emotional clarity, difficulties engaging in goal-directed behavior, impulse control difficulties, non-acceptance of emotional responses, and limited access to effective emotion regulation strategies.

Furthermore, there may be a reciprocal relationship between depression and DER. Individuals with depression often have trouble with preventing the detailed processing of negative emotions, which may lead to the use of more non-adaptive emotion regulation strategies and greater difficulties in managing their emotions [20]. On the other hand, specific dimensions of difficulty in regulating emotions may contribute to depressive symptoms. For instance, emotional clarity–a facet of emotional awareness‒refers to the ability to identify, understand, and discriminate among emotional experiences [21]. Individuals who struggle to recognize and distinguish their emotions (e.g., feelings of depression vs. anger) are less likely to choose effective regulation strategies, thereby enhancing their susceptibility to depression [22]. Conversely, higher levels of emotional clarity may help reduce participants’ levels of depression [23]. Additionally, non-acceptance manifests as having secondary emotional reactions (e.g., shame) to primary emotions or non-acceptance responses to one’s suffering [24]. Those with high levels of non-acceptance tend to avoid or suppress negative emotions and are less likely to adopt adaptive strategies, which may elevate the risk of depression [24]. Empirical research has shown that greater non-acceptance of emotional responses predicted higher levels of depression [25]. While there is evidence that emotional regulation difficulties and depression are comorbid, most studies have overlooked their interactions at the symptom level.

Depression and DER are typically measured using total scores calculated from ratings on items of standardized scales [19, 26]. However, utilizing total scores obscures heterogeneity in individual symptoms when assessing DER or depression. Additionally, this approach does not consider symptom interactions between the two disorders. Network theory offers new insights into symptom interactions across mental disorders [27, 28]. The network theory posits that mental disorders emerge from dynamic symptom interactions within a complex network [29, 30]. Specific symptoms can activate interrelated symptoms, and symptom interactions strengthen and sustain the network [29, 30]. Network analysis, based on network theory, explains symptom interactions and models symptom networks for certain disorders [29]. It generates a network graph with nodes and edges [28]. The nodes signify symptoms and the edges represent their associations. Bridge symptoms connect the comorbid disorders, and are considered to be an underlying mechanism that triggers or maintains comorbidities [31].

Prior network studies have identified robust connections between indicators of depression and difficulties in emotion regulation [32–34]. For example, Ruan et al. [32] used network analysis to examine associations among nodes of emotion regulation, emotional reactivity, depression, and anxiety. They found that “non-acceptance of emotional responses” and “limited access to effective emotion regulation strategies” were linked to “depression” via “lack of emotional clarity”. However, depression was treated as a single aggregated node, rather than as discrete symptoms in the study. Liang et al. [33] analyzed the network of individual depressive symptoms and two emotion regulation dimensions: cognitive reappraisal and expressive suppression. They identified “thoughts of death” and “expressive suppression” as bridge symptoms. Yet, this study was limited by its focus on two DER dimensions and did not comprehensively capture interactions between depression and DER. A more recent study [34] explored interrelationships between depressive symptoms and DER dimensions, and identified “limited access to effective emotion regulation strategies” and “I feel cheerful” as key bridge nodes. However, this study focused exclusively on adolescents seeking clinical aid and had a small sample size (n = 209). Therefore, network analysis is needed to identify the bridge symptoms between depression and DER among first-year college students, potentially informing targeted treatments.

The recognition of emotional clarity is considered as an essential first step in effective emotion regulation [35, 36]. Emotional comprehension assists the allocation of cognitive resources for subsequent emotion regulation [37], because effective regulation can only be achieved if one knows what needs to be modified. The emotional ambiguity may impede effective emotion regulation, contributing to disorders like depression [38, 39]. In addition, emotional non-acceptance plays a significant role in emotional problems [40]. Extensive research has consistently demonstrated a strong correlation between high levels of non-acceptance and various forms of psychopathology, including depression [41–43]. Hence, emotional clarity and non-acceptance could be important bridge symptoms between DER and depression.

To our knowledge, there has been a lack of utilization of network methods to investigate the correlation between depression and DER within a sample of first-year college students. The present study used network analysis to examine the depression-DER network structure among first-year college students, and to identify the bridge symptoms in the network. It is hypothesized that emotional clarity and non-acceptance could be central bridge symptoms.

Methods

Participants and procedure

A total of 1050 participants who were from four universities located in Zhejiang province, China, participated in this study. They were recruited via flyers, posters, and direct contact with instructors. Participation was voluntary with no incentives offered. The inclusion criteria were as follows: (1) first-year college students who could comprehend the objectives and procedure of the evaluations; (2) students whose native language is Chinese. The exclusion criteria were (1) participants who fail to fully complete or provide incomplete responses to the questionnaires; and (2) participants who did not adhere to a reading check question (i.e., please select the strongly agree option). A total of 992 valid questionnaires (M_age = 18.68, SD = 0.85, 64.31% females, 93.3% Han nationality) were received, with the effective response rate of 94.48%.

Ethical approval was obtained from the ethics committee of Zhejiang Normal University (No. ZSRT2024136). Following approval from the relevant administrations of universities, posters about the survey were displayed on designated campus bulletin boards in high-traffic areas such as student centers and departmental buildings. At the same time, flyers were distributed to further promote the survey among students. Participants completed self-administered, paper-and-pencil questionnaires at private workstations in separate, quiet classrooms. To ensure participant privacy, multiple safeguards were implemented: (1) anonymous response collection with no personal identifiers; and (2) adequate physical spacing between participants. A standardized script, read aloud by a trained research assistant, explained the study’s purpose, voluntary participation, confidentiality, and the right to withdraw without penalty. All participants provided their written informed consent, and for minors, guardian assent was obtained by mail.

Measures

Patient health questionnaire (PHQ-9)

The Chinese version of the 9-item Patient Health Questionnaire (PHQ-9) [26, 44] was utilized to measure depressive symptoms. It reflects the nine diagnostic criteria for depression outlined in the Diagnostic and Statistical Manual of Mental Disorders, 5th edition (DSM-5) [45]. The nine items assess anhedonia, depressed mood, sleep problems, fatigue, appetite problems, self-blame, poor concentration, motor problems, and suicidal ideation. Participants rated the frequency of symptoms experienced over the past two weeks. Items were rated on a 4-point scale ranging from 0 (not at all) to 3 (almost daily). Higher scores indicated a more severe manifestation of depressive symptoms. Previous studies have exhibited the PHQ-9’s measurement invariance between clinical and non-clinical populations [46, 47]. Its reliability and validity have been well established in Chinese college students [44, 47]– [48]. In the present study, the Cronbach’s α was 0.83.

Difficulties in emotion regulation scale (DERS-16)

The brief version of Difficulties in Emotion Regulation Scale (DERS-16) [19, 49] was used to assess DER. The scale included 16 items, which were loaded on 5 dimensions: lack of emotional clarity, difficulties engaging in goal-directed behavior, impulse control difficulties, non-acceptance of emotional responses, and limited access to effective emotion regulation strategies. Items were rated on a 5-point scale ranging from 1 (hardly ever) to 5 (always). Higher scores indicated greater difficulties in emotion regulation. The scale demonstrated good reliability and validity in Chinese college students [50–52]. In this study, the Cronbach’s α was 0.92.

Data analysis

Descriptive analyses of variables were conducted using SPSS 26.0. Network analysis was carried out in R version 4.1.2.

Network estimation and visualization

Networks were estimated and visualized using the R-package qgraph [53]. The estimated network was a Graphical Gaussian Model (GGM) (network of partial correlation coefficients). The GGM uses LASSO (Least Absolute Shrinkage and Selection Operator) regularization to reduce weak edges to zero, resulting in a sparser network. This process is controlled by adjusting a tuning hyperparameter (λ), which ranges from 0 (less conservative) and 1 (more conservative). In this study, λ was set to 0.5 to balance network sensitivity and specificity [54]. In the visual network, nodes represent symptoms, and edges represent partial correlations between them. Thicker edges signify stronger correlations between symptoms.

Bridge nodes

This study conceptualized two communities a priori: depression (nine symptoms) and difficulties in emotion regulation (five symptoms). The R-package networktools was used to identify bridge nodes by computing bridge expected influence (BEI) [55]. The BEI is defined as the sum of edge weight that connects a node in one community to all nodes in another community. Higher BEI indicates a greater probability of activating relevant communities.

Edge weight accuracy and centrality stability

Since estimated networks and BEI indices may deviate from true structures and BEI values when sample sizes vary [55], we evaluated edge weight accuracy and BEI stability using the R-package bootnet [56]. Edge weight accuracy was assessed via non-parametric bootstrapping with 2,000 resampled datasets to compute 95% confidence intervals (CIs). Narrower CIs indicate greater estimation precision and stability. BEI stability was evaluated using the correlation stability (CS) coefficient, estimated through a case-dropping bootstrap procedure with 2,000 samples. This coefficient reflects the maximum proportion of cases that can be removed while maintaining a strong correlation (> 0.70) between the original and subset bridge centrality indices with a 95% probability [56]. A CS coefficient not lower than 0.25 is acceptable, with a preference for values exceeding 0.5 [56]. Additionally, the difference test function (with 2000 bootstrap samples) in the R package bootnet was used to test for differences between the BEIs of nodes and between individual edge weights [56].

Network comparison tests

To check if male and female networks were different, we compared structure invariance, global strength, and edge invariance via the R package NetworkComparisonTest (NCT) [57]. The NCT uses non-parametric permutation tests (random permutations, n = 1000). Structure invariance examines the variations in the distributions of edge weights across different networks. Global strength refers to the sum of all edge weights. Edge invariance is the difference of each edge weight between two networks.

Results

Descriptive statistics

Table 1 shows abbreviations, mean scores and standard deviations for each variable selected in the depression-DER networks.

Table 1.

Node abbreviation and description of depression scale and difficulties in emotion regulation scale

Node name	Item content	Mean	SD
PHQ1	anhedonia	1.06	0.64
PHQ2	depressed mood	0.88	0.63
PHQ3	sleep problems	0.72	0.81
PHQ4	fatigue	1.20	0.74
PHQ5	appetite changes	0.84	0.81
PHQ6	self-blame	0.86	0.77
PHQ7	concentration difficulties	0.89	0.83
PHQ8	motor problems	0.43	0.63
PHQ9	suicidal ideation	0.22	0.52
DER1	lack of emotional clarity	2.13	0.85
DER2	difficulties engaging in goal-directed behavior	3.06	0.98
DER3	impulse control difficulties	2.12	0.96
DER4	non-acceptance of emotional responses	2.27	0.84
DER5	limited access to effective emotion regulation strategies	2.22	0.90
	Depression scale	7.10	4.17
	DER scale	37.81	11.60

Note. PHQ, Patient Health Questionnaire; DER, Difficulties in Emotion Regulation

Network estimation

Figure 1 shows a network comprising depressive symptoms and DER dimensions. In the current network, a total of 68 out of the 91 potential edges were set to non-zero edge weights. All the edges were positive, indicating positive partial correlations between the symptoms. Three strongest edges were between DER4 (non-acceptance of emotional reactions)-DER5 (limited access to effective emotion regulation strategies) (edge weight = 0.45), DER4 (non-acceptance of emotional reactions)-DER3 (impulse control difficulties) (edge weight = 0.38), and DER4 (non-acceptance of emotional reactions)-DER2 (difficulties engaging in goal-directed behavior) (edge weight = 0.34). The strongest and weakest edges were significantly different from one another (see Supplementary Table S1 and Figure S1).

Fig. 1.

Network structure of depression-DER. Note. Depressive symptoms are represented by the color deepskyblue, while DER symptoms are represented by palevioletred. Each node in the network is represented by a circle in the diagram. The thickness of the edges represents the intensity of the connection between nodes. Nodes with a bridge expected influence (BEI) value in the top 15% of all nodes are classified as belonging to the bridge group and are displayed in white

Bridge centrality

Figure 2 shows the value of the BEI for each symptom. The white nodes in Fig. 2 represent the two nodes with BEI to be in the top 15% (DER1: lack of emotional clarity [BEI = 0.33] and DER4: non-acceptance of emotional responses [BEI = 0.33]). Bootstrapping difference tests showed that these two nodes were significantly stronger than all other nodes (see Supplementary Figure S2). Thus, lack of emotional clarity and non-acceptance of emotional responses were crucial nodes linking DER to depression.

Fig. 2.

Bridge expected influence (BEI) value of each variable in the network

Network accuracy and stability

Most (non-zero) edges showed considerable overlap at 95% CIs, suggesting that edge weights are accurate (see Supplementary Figure S3). The CS-coefficient for BEI (CS [cor = 0.7] = 0.69) was greater than 0.25, indicating that the bridge centrality estimates were stable (see Supplementary Figure S4).

Comparison of male and female networks

The depression-DER network diagrams for females and males are provided in the Figs. 3 and 4. There were no significant differences in network structure (maximum difference = 0.15, p = 0.47) or global strength (females = 5.83, males = 5.69, global strength difference = 0.13, p = 0.65) between females and males. However, some edge weights exhibited differences: PHQ5 (appetite changes)-PHQ6 (self-blame) (E = 0.11, p = 0.042), PHQ2 (depressed mood)-DER4 (non-acceptance of emotional responses) (E = 0.11, p = 0.034), and PHQ4 (fatigue)-PHQ6 (self-blame) (E = 0.12, p = 0.005) were greater in males; DER1 (lack of emotional clarity)-DER3 (impulse control difficulties) were greater in females (E = 0.15, p = 0.011). No other edges differed significantly (all p > 0.05).

Fig. 3.

Comparison of network structure between females (left panel) and males (right panel)

Fig. 4.

Comparison of network properties between females and males. Note: Left panel: A plot of bootstrap value of the maximum difference in any of the edge weights (1000 permutations). The difference was not significant (maximum difference = 0.15, p = 0.47). Right panel: A plot of bootstrap value of the difference in network global strength. The difference was not significant (females = 5.83, males = 5.69, global strength difference = 0.13, p = 0.65)

Discussion

This study used network analysis to examine connections between depression and DER symptoms among first-year Chinese college students. The estimated depression-DER network demonstrated high accuracy and good stability in BEI. Two key bridge symptoms connecting depression to DER were identified: “lack of emotional clarity (DER1)” and “non-acceptance of emotional responses (DER4)”.

The present study revealed that “lack of emotional clarity” was a key bridge symptom linking DER and depression among first-year college students, thereby supporting our hypothesis. This finding aligns with prior research highlighting the central role of emotional clarity [58–60]. For instance, Ruan et al. [32] demonstrated that “lack of emotional clarity” acted as a critical bridge node linking “non-acceptance of emotional responses” and “limited access to effective emotion regulation strategies” to depression. Accurately perceiving and understanding emotions constitutes an initial stage of emotion regulation [36]. Individuals with difficulty in understanding and distinguishing their emotions (i.e., low emotional clarity) may be less likely to implement adaptive emotion regulation strategies, thereby contributing to the emergence and persistence of mental disorders such as depression [27, 38, 61]. In addition, individuals with low emotional clarity may exhibit a tendency to prioritize the comprehension of the feelings elicited by a stressor, rather than actively seeking solutions to address the underlying problems [59, 62]. Given that first-year college students face numerous pressures [2–4], lack of emotional clarity may function as a potential mechanism underlying emotional problems (including depression) that arise from these pressures. To sum up, lack of emotional clarity plays a significant role in the development and maintenance of the depression-DER symptom network.

“Non-acceptance of emotional responses” also emerged as a critical bridge symptom. This finding supports previous studies indicating that “non-acceptance of emotional responses” mediated the path between fibromyalgia and depression [25], and that the capacity to adopt a mindset of acceptance and openness towards negative emotions serves as a preventative measure in the emergence of psychopathology [12, 63]. Notably, however, this result contrasts with a prior study [34], which found that “limited access to effective emotion regulation strategies,” rather than “non-acceptance of emotional responses,” was the key bridge symptom. This discrepancy may stem from differences in network focus (depression-DER network vs. depression-anxiety-DER network), measurement instruments (PHQ-9 vs. Hospital Anxiety and Depression Scale), and participant characteristics (college students vs. adolescents seeking for clinical treatments for depression). In sum, the present finding suggests that non-acceptance or avoidance of negative emotions may exacerbate depression by limiting the use of effective strategies and prolonging maladaptive affective states in first-year college students. Conversely, accepting emotions might facilitate adaptive regulation, enabling students to respond constructively to pressures and effectively utilize their emotional experiences in pursuit of personally significant actions [64].

In the depression-DER network, the three strongest positive edges were found between “non-acceptance of emotional responses (DER4)” and “limited access to effective emotion regulation strategies (DER5)”, “impulse control difficulties (DER3)”, and ‘difficulties engaging in goal-directed behavior (DER2)”, respectively. These patterns emphasize the importance of emotional acceptance in connecting various symptoms. These findings are consistent with previous studies, which showed that individuals who did not acknowledge negative emotions were more likely to use maladaptive emotion regulation strategies, such as rumination [65] and inhibition [66], and experienced difficulties in controlling their behavior [18]. These results suggest that it is crucial to help first-year college students better understand and accept negative emotions.

Overall, female and male college students showed similar network structure and global strength. However, males exhibited higher edge weights than females on three links, and the edge with the greatest difference was “fatigue (PHQ4)” - “self-blame (PHQ6)”. The possible explanation could be grounded in Chinese cultural norms, which emphasize that men should manifest heightened toughness. Consequently, when they feel fatigue, they may perceive a failure to meet societal expectations imposed upon males and tend to blame themselves [67]. In addition, it is noteworthy that the link between “depressed mood (PHQ2)” and “non-acceptance of emotional responses (DER4)” was also stronger in male students. One possible explanation is that masculinity can bring out the fear of vulnerable emotions in boys. Thus, masculinity norms may make acceptance of depressive mood harder for male Chinese college students [68, 69]. In conclusion, these results provide new perspectives on the mechanisms underlying the coexistence of depression and DER across genders.

The present findings have theoretical and clinical implications. To our knowledge, this is the first study that employed network analysis to examine the interrelationships between depressive symptoms and DER components among first-year college students. Therefore, this study offers a potential foundation for understanding comorbidity between depression and DER, supplementing extant literature [20, 22–24, 33]. In addition, the present study identified “lack of emotional clarity” and “non-acceptance of emotional responses” as core bridge symptoms. Researchers propose that activating bridge symptoms leads to activation of symptoms within other disorders [70]. Thus, targeting bridge symptoms may help prevent the co-occurrence of depression and DER. Mental health practitioners can help college students or the general public understand that enhancing emotional clarity and strengthening self-acceptance of emotions may be more effective in alleviating depressive symptoms, and they can intervene the two symptoms through targeted interventions in their clinical practice.

Specifically, mindfulness exercises were shown to facilitate the practice of observing one’s experiences and cultivating awareness of those experiences, increasing emotional clarity [23, 71]. On the other hand, acceptance of emotional responses can be addressed by acceptance commitment therapy (ACT), which enables participants to actively confront and disengage from negative feelings [72]. The school counseling center could offer mindfulness exercises and ACT group courses, available both online and as part of the general education curriculum. Online courses could also be extended to the general public. Additionally, the emotion recognition training (ERT), a computer-based program, enables participants to repeatedly practice recognizing facial expressions, micro-expressions, vocal features, thereby gradually enhancing an individual’s sensitivity and accuracy in identifying complex emotional information [73]. Mental health practitioners can use these programs to help college students or the general public to improve their emotional clarity and alleviate depressive symptoms [74].

This study had some limitations. First, the data were obtained from a cross-sectional survey and cannot establish causality between nodes. Future longitudinal studies should investigate the evolving nature of symptoms over time. Second, this study relied on self-report scales and may introduce recall bias. Future research could incorporate behavioral observations to provide a more objective assessment of emotion regulation. Third, this study employs the DERS-16 instead of the full 36-item version (DERS-36), which also include the “lack of emotional awareness” component. While previous research has reported overlaps between the “lack of emotional awareness” dimension and other dimensions (e.g., “limited access to emotion regulation strategies” and “impulse control difficulties”), and the DERS-16 has been increasingly used in recent years, utilizing the DERS-36 to measure emotion regulation difficulties will provide a more comprehensive assessment and potentially extended the findings of this study. Finally, the sample of this study was limited to the first-year college students from a single province in China, limiting generalizability of the results to other regions. Future studies should use a more diverse sample of first-year college students to validate and replicate the current results.

Conclusion

This study conducted a network analysis on first-year college students to examine the intricate interactions between depression and DER symptoms. The findings were as follows: (1) “Lack of emotional clarity” and “non-acceptance of emotional responses” emerged as most important bridge symptoms that connected depression and DER communities; (2) The strongest connections within depression-DER network were found between “non-acceptance of emotional reactions” and “limited access to effective emotion regulation strategies”, “impulse control difficulties” and “difficulties engaging in goal-directed behavior”, respectively; and (3) The network structures and connectivity were largely similar across genders. According to our results, using tailored interventions targeting the key bridge nodes might hinder the development of depression-DER comorbidity among college students during the transitioning period.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

LL: Conceptualization, Methodology, Data curation, Formal analysis, Writing - Original Draft, Writing - Review & Editing. TS: Conceptualization, Methodology, Data curation, Visualization, Writing - Original Draft. JC: Conceptualization, Supervision, Writing - Review & Editing. YC: Writing - Review & Editing. GL: Writing - Review & Editing.LL and TS served as co-first authors.

Funding

This research was supported by the China National Social Science Fund in Education (2023 General Project. Grant No. BBA230058).

Data availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the guidelines of the Declaration of Helsinki. This study was reviewed and approved by the ethics committee of Zhejiang Normal University (No. ZSRT2024136). All participants provided their written informed consent, and legal guardian assent was requested by mail.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.