Factors influencing suicidal ideation in Chinese adolescents with first-episode depressive disorder: a cross-sectional study

Abstract

Background

Given the high prevalence and severe consequences of suicidal ideation (SI) in adolescents, it is critical to identify multidimensional predictors of the illness. This study aimed to investigate the physiological, psychological, and sociological factors that influence SI in adolescents with first-episode depressive disorder.

Methods

The study was recruited through convenience sampling. Data were collected using self-designed questionnaires, the Hamilton Depression Scale 24-item, the Hamilton Anxiety Scale, and the Simplified Coping Style Questionnaire. Thyroid function, cortisol, lipids, and event-related potential were measured in the participants. All independent variables were included in the model for logistic regression. Statistical analysis of data was done using SPSS version 25.0.

Results

The present study was for the inclusion of 150 adolescents who presented with first-episode depressive disorder, of whom 96 (64.00%) had SI. Females and older adolescents were more prone to SI. Depressive symptoms, anxiety symptoms, negative coping styles, total cholesterol levels, and the latency of P3a and P3b were positively correlated with SI. Positive coping style, N2 amplitude, and SI were negatively correlated.

Conclusions

SI in Chinese adolescents with first-episode depressive disorder demonstrates a multifactorial relationship, involving physiological, psychological, and sociological factors. It is necessary to conduct multidomain, comprehensive assessments and develop intervention strategies.

Introduction

Adolescence is a critical stage of transition from childhood to adulthood [1]. During this period, individuals often experience heightened emotional volatility, increased sensitivity to social evaluation, and difficulty establishing a stable self-identity [2, 3]. Erikson’s psychosocial development theory conceptualizes this stage as ‘identity vs. role confusion,’ which is marked by a core conflict between forming a coherent self-identity and succumbing to social pressures and expectations [4]. Bronfenbrenner’s bioecological model further clarifies adolescent vulnerability by highlighting the influence of interactions across environmental systems on development, ranging from the immediate microsystem (e.g., family, peers) to the broader macrosystem (e.g., cultural norms) [5]. Within these systems, adolescents face intensified pressures, which can increase their susceptibility to mental health issues [6, 7].

Among these mental health issues, adolescent depressive disorder and suicidal ideation (SI) have emerged as major global public health concerns. The overall prevalence of SI among adolescents in low- and middle-income countries is approximately 16.9% [8]. SI frequently recurs during depressive episodes, increasing the risk of recurrent depressive disorder. Over 58% of major depressive disorder (MDD) patients have experienced SI, a known precursor to suicidal behavior (SB) [9].

Within the complex etiology of SI, psychosocial factors like low self-esteem and trauma history are established risks [10, 11]. Among these, coping style may serve as one key psychological mechanism through which stress is translated into SI, which comprises the cognitive and behavioral strategies used to manage stress [12]. Because adolescents’ still-maturing prefrontal cortex limits their capacity for adaptive, top-down emotion regulation, they increase reliance on a negative coping style [13, 14]. Studies on university students from China show that coping through guidance-seeking and problem-solving positive coping style reduces SI risk, whereas habitual submission or emotional venting negative coping style raises it [15]. The appraisal of a stressful event and the subsequent coping response are critical determinants of psychological outcomes [16].

However, SI diagnosis in adolescents, including assessment of psychological factors like coping style, relies mostly on scales and clinical interviews. A major limitation of these methods is dependence on subjective reporting. Adolescents may underreport or deny SI due to fluctuating symptoms, fear of stigma, or concerns about triggering involuntary interventions [17–19]. Poor insight into their own mental state and the role of implicit cognitive processes can also reduce reporting accuracy [20, 21]. This makes it difficult to rely exclusively on subjective measures for a valid assessment [22–24]. Exploring complementary, objective risk assessment avenues for SI is necessary.

Most research on objective biological markers has been in adults, offering a preliminary basis for cautious extension to adolescents. Biomarkers reflecting metabolic function may be part of SI and SB-related pathophysiology. In adults, elevated levels of thyroid-stimulating hormone (TSH) and certain thyroid antibodies have been observed in patients with SI compared to those without SI [25]. Dysregulation of the hypothalamic-pituitary-thyroid axis has also been proposed as a potential risk factor for SB in depressive disorder [26]. Similarly, lipid profiles are not merely cardiovascular risk markers but are also involved in brain structure and function, including neuronal membrane integrity and serotonin receptor activity [27–29]. Shen et al. found a negative correlation between SI and low TG levels [30]. Emerging evidence in adolescents points to an elevated total cholesterol (TC)/high-density lipoprotein (HDL) ratio has been significantly associated with a higher likelihood of SI [31]. During adolescence, a period of high metabolic demand for synaptic pruning and myelination, alterations in these metabolites could disrupt neural plasticity and neurotransmission, thereby influencing impulsivity and depressive symptoms that underlie SI [32].

Neurophysiological approaches have also emerged as a promising avenue for understanding SI [33]. Event-related potential (ERP) refers to brain potentials evoked by stimuli with psychological significance. Classic ERP components include exogenous ones (P1, N1, P2), related to physical characteristics of the objective stimuli, and endogenous ones (N2, P3), which reflect activities related to the individual’s psychological changes [34]. N2 represents a transitional component between automatic processing and controlled processing. It is a negative potential that usually appears peaking approximately 200 ms after the stimulus is presented [35]. P3 is a positive potential that appears peaking approximately 300 ms after an event-related stimulus [36]. Specifically, P3 has two subcomponents. P3a often reflects the perception of conflicting stimuli and the initial response to novel stimuli [37]. P3b is often used to reflect processes such as information extraction and recognition in individuals [37]. The ERP latency or amplitude abnormality may represent the core psychological profile of suicidality: difficulties in disengaging from salient, negative internal cues, impaired problem-solving under distress, and a predisposition toward impulsive decisions [38, 39]. Previous adult results suggest that adults with depressive disorder who have attempted suicide have reduced central serotonin function and elevated frontal P3 amplitude [40]; adolescents with MDD who have non-suicidal self-injury (NSSI) have a prolonged latency and decreased amplitude of P3, as well as cognitive impairments such as executive dysfunction and memory impairment [41].

Given that adolescents exhibit distinct emotional and developmental vulnerabilities compared to adults, research specifically targeting this population is essential. Meanwhile, investigating first-episode depressive disorder is crucial, as it minimizes confounding from chronic illness duration, repeated medication use, and the accumulated psychosocial burden of recurrent episodes, thereby offering a clearer window into the core mechanisms underlying SI onset. Moreover, early identification of risk at this stage holds strong preventive potential. Therefore, this study aims to construct an integrated risk model spanning physiological, psychological, and social dimensions. This model will be used to compare adolescents with first-episode depressive disorder with and without SI, identify key predictors of SI, and ultimately inform better diagnostic and intervention strategies for this population.

Methods

Participants

The study was a cross-sectional survey conducted from January to October 2022, using convenience sampling. Participants were adolescents admitted to Daizhuang Hospital in Shandong Province. Two attending psychiatrists independently evaluated participants according to the International Statistical Classification of Diseases and Related Health Problems 10th Revision (ICD-10) diagnostic criteria. Eligibility criteria were as follows: (1) aged 10–19 years, male or female; (2) meeting ICD-10 diagnostic criteria for first-episode depressive disorder, with no prior medication, psychotherapy, or physical therapy within the 3 months before enrollment; (3) Hamilton Depression Scale 24-item (HAMD-24) total score ≥ 20; (4) understanding the study purpose and content and willing to cooperate; and (5) informed consent obtained from both patients and their guardians.

The epidemiological sample size used in this study was calculated using the following formula [42]: N = (Z_α² × P × (1–P))/d². According to a previous study, the prevalence of SI among Chinese adolescents in grades 5–12 was approximately 13%. An estimated proportion of P = 0.13 [43], a significance level of α = 0.05, a confidence level of Z = 1.96, and a permissible error of d = 0.10 were set. A sample size of at least 43 was thus determined.

Measurements

Socio-demographic information: A self-designed questionnaire was used to collect data on gender, age, place of residence, duration of illness, etc., from all participants.

HAMD-24: This scale was used to assess depressive symptoms over the past week [44]. 13 items are rated on a 5-point scale (0–4), and 11 on a 3-point scale (0–2). The total score on the scale ranges from 0 to 74, with higher scores indicating more severe depressive symptoms. SI was assessed using item 3 of the scale, namely the Hamilton Depression Scale Suicide-Item (HAMD-SI), where a score of 0 indicates the absence of SI. SI was considered present under other scoring conditions, such as “1 score: Feels life is not worth living”, “2 score: Wishing to be dead, or frequently thinking about death”, “3 score: Negative thoughts”, or “4 score: Serious SB”. The Cronbach’s alpha coefficient for the Chinese version of the HAMD-24 in this study was 0.903.

Hamilton Anxiety Scale (HAMA): Compiled by Max Hamilton in 1959, this scale was used to assess the severity of anxiety symptoms over the past week [45]. The Chinese version of the scale has demonstrated good reliability and validity in epidemiological studies involving Chinese adolescents [46]. The scale comprises 14 items rated on a 5-point Likert scale ranging from 0 to 4, resulting in a total score ranging from 0 to 56, with higher scores indicating more severe anxiety symptoms. In this study, the Cronbach’s alpha coefficient for the Chinese version of the scale was 0.912.

Simplified Coping Style Questionnaire (SCSQ): The SCSQ is a 20-item self-report scale designed to assess an individual’s coping style [47]. It consists of two factors: positive coping (12 items) and negative coping (8 items). Scores on the SCSQ indicate participants’ preferences in coping style, with the positive coping factor reflecting active coping style, such as “Finding several different solutions when faced with a problem”. In contrast, the negative coping factor assesses a passive coping style, for example, “Escaping from problems by drinking and smoking”. Each item is rated on a 4-point Likert scale, where higher scores on each factor indicate a stronger inclination toward that coping style. In this study, the Cronbach’s alpha coefficient for the scale was 0.869.

Blood samples

After enrollment, all participants underwent an overnight fast, and blood samples were collected from the anterior elbow vein the following day (6:30 AM to 8:00 AM). The samples were processed in the hospital laboratory department by placing peripheral blood into anticoagulant-containing blood specimen collection tubes for 30 min, followed by centrifugation at 3000 rpm for 15 min. TSH, free triiodothyronine (FT3), and free thyroxine (FT4) levels were measured in the serum samples using electrochemiluminescence immunoassay on an Autolumo A2000 fully automated chemiluminescence immunoassay analyzer (Autobio, Jinan, China). TC, triglycerides (TG), low-density lipoprotein (LDL), HDL, and cortisol (COR) data were also collected.

ERP detection

ERP data were collected using the NVX36 ERP-electroencephalography (EEG) integrated machine (Biaosi, Shenzhen, China). Participants sat in a quiet and soundproof laboratory, maintaining emotional stability. To eliminate artifacts caused by activities such as blinking, participants were asked to close their eyes, relax their entire body, but remain awake. Ag/AgCl disc electrodes were used, placed according to the international 10/20 standard EEG system. Before electrode placement, the scalp was thoroughly cleaned of oils, and conductive gel was applied to the reference electrodes. Electrode positioning and fixation with adhesive tape were carried out to ensure that the electrode-skin contact resistance was less than 5 kΩ. Recording electrodes were placed at frontal (Fz), central (Cz), and parietal (Pz) positions, with reference electrodes placed at the left and right mastoids (A1 and A2), and the ground electrode at the fronto-central (FCz) point. The EEG data were processed through a 0.1–100 Hz band-pass filter and collected at a sampling rate of 1000 Hz.

The classic oddball paradigm was used for the Go/No-Go task. Sounds were presented binaurally in the form of pure tones, with a series of high-frequency (80%) standard stimuli (1000 Hz audio, 60 dB), among which low-frequency (20%) deviant stimuli (2000 Hz audio, 80 dB) were randomly inserted. Participants were instructed to press a button immediately upon hearing the deviant stimulus, but not to respond to standard stimuli. A total of 300 sounds were presented, with 240 standard stimuli and 60 deviant stimuli. EEG data were extracted from the period 500 ms before the stimulus to 900 ms after the stimulus, and artifacts with amplitudes exceeding ± 75 µV were removed. Waveform identification used the Cz electrode as the standard. The latency and amplitude of N2 (the maximum negative potential between 150 ms and 250 ms after stimulus onset) and P3 (the maximum positive potential between 300 ms and 450 ms after stimulus onset) were quantified at the Cz electrode. The requirement to respond “Go” to low-frequency deviant stimuli while withholding responses “No-Go” to high-frequency standard stimuli places demands on response inhibition and conflict monitoring, which are reliably reflected in the N2 and P3 components. This makes the paradigm suitable for investigating the neurocognitive correlates of SI, as deficits in these processes may underlie difficulties with impulse regulation and disengagement from negative stimuli [48].

Data analysis

SPSS 25.0 was used for statistical analysis of data in this study. Participants were categorized into two groups: the SI group and the No-Suicide Ideation (NSI) group. Continuous variables were expressed as mean ± standard deviation. Categorical variables were presented as frequencies (%). Between-group differences in categorical variables (gender, age, and place of residence) were assessed using the chi-square test. Continuous variables were compared between groups using either the two independent samples t-test or the Mann-Whitney U-test, depending on normality verification through histograms or the Shapiro-Wilk test. All independent variables involved in the study were included in the logistic regression model in ascending order to assess the influence of physiological, psychological, and sociological factors on SI in adolescents with first-episode depressive disorder. All statistical tests were two-tailed, and a P value of less than 0.05 was considered statistically significant.

Results

Prevalence of SI and demographic/psychological characteristics

A total of 150 participants were enrolled in this study, comprising 76 males (50.67%). Among them, 96 individuals (64.00%) belonged to the SI group and 54 (36.00%) to the NSI group. Within the SI group, the HAMD-SI item assessing SI revealed a distribution of severity: 33 participants felt life was not worth living (score = 1), 31 reported wishing to be dead, or frequently thinking about death (score = 2), 18 had negative thoughts (score = 3), and 14 demonstrated serious SB (score = 4). The following continuous variables were normally distributed: FT3, TG, N2 latency, P3a latency, N2 amplitude, P3a amplitude, and P3b amplitude; all other continuous variables deviated from normality. As presented in Table 1, statistically significant differences were identified between the SI and NSI groups in positive coping scores (P < 0.001), negative coping scores (P < 0.001), HAMA scores (P < 0.001), and HAMD-24 scores (P < 0.001). The SI group exhibited more severe anxiety and depressive symptoms, together with a greater tendency toward negative coping strategies.

Table 1.

The univariate analysis of demographic and psychological variables in adolescents with SI or NSI

Variables	SI (N = 96)	NSI (N = 54)	χ2/Z	P
Gender			0.015	0.903a
Male	49 (51.04)	27 (50.00)
Female	47 (48.96)	27 (50.00)
Residence			0.047	0.828a
Rural	48 (50.00)	28 (51.85)
Urban	48 (50.00)	26 (48.15)
HAMD-SI
1 score (Feels life is not worth living)	33 (34.38)	-	-	-
2 score (Wishing to be dead, or frequently thinking about death)	31 (32.29)	-	-	-
3 score (Negative thoughts)	18 (18.75)	-	-	-
4 score (Serious suicidal behaviour)	14 (14.58)	-	-	-
Age (years)	15.03 ± 1.71	14.61 ± 1.88	−1.315	0.188b
Age of onset (years)	12.33 ± 1.55	11.94 ± 1.81	−1.723	0.085b
Disease duration (years)	2.70 ± 1.35	2.67 ± 1.58	−0.511	0.609b
Educational attainment (years)	7.96 ± 1.67	7.46 ± 1.79	−1.708	0.088b
HAMA	17.99 ± 4.07	12.33 ± 3.25	−7.089***	< 0.001b
HAMD-24	28.25 ± 5.57	23.70 ± 2.90	−5.277***	< 0.001b
Positive Coping Style	26.28 ± 6.53	33.04 ± 5.05	−6.202***	< 0.001b
Negative Coping Style	22.72 ± 4.14	16.44 ± 3.01	−7.801***	< 0.001b

SI Suicidal Ideation, NSI No-Suicide Ideation, HAMD-SI Hamilton Depression Scale Suicide-Item, HAMA Hamilton Anxiety Scale, HAMD-24 Hamilton Depression Scale 24-item

Values are presented as mean ± standard deviation or frequencies (%)

^aChi-square test

^bMann-Whitney U-test

^*P < 0.05, ^**P < 0.01, ^***P < 0.001

Biochemical and ERP parameters of the SI group vs. the NSI group

As shown in Table 2, the SI group had significantly higher levels of TC (P < 0.001), TG (P < 0.05), and LDL (P < 0.01) than the NSI group. Adverse lipid profiles may serve as potential peripheral biomarkers associated with increased SI, highlighting the value of incorporating metabolic screening into comprehensive risk assessments. In terms of ERP, the SI group showed significantly longer P3b latency (P < 0.001), suggesting slower stimulus evaluation and slower cognitive processing. In contrast, N2 amplitude was significantly lower in the SI group (P < 0.01). This decrease in N2 amplitude is associated with potential deficits in cognitive processing.

Table 2.

The univariate analysis of biochemical and ERP parameters in adolescents with SI or NSI

Variables	SI (N = 96)	NSI (N = 54)	Z/t	P
FT3 (mIU/L)	4.99 ± 0.80	5.04 ± 0.76	0.404	0.687a
FT4 (pmol/L)	14.84 ± 2.01	14.94 ± 1.68	−0.018	0.986b
TSH (pmol/L)	2.43 ± 0.96	2.71 ± 0.91	−1.709	0.087b
COR (nmol/L)	328.35 ± 120.96	335.80 ± 149.82	−0.484	0.629b
TC (mmol/L)	1.54 ± 0.43	1.40 ± 0.41	−3.532***	< 0.001b
TG (mmol/L)	1.54 ± 0.43	1.40 ± 0.41	−2.623**	0.010a
LDL (mmol/L)	2.43 ± 0.56	2.14 ± 0.51	−2.992**	0.003b
HDL (mmol/L)	1.32 ± 0.17	1.13 ± 0.13	−0.831	0.406b
ERP latency (ms)
N2 latency	216.35 ± 27.75	216.61 ± 29.79	0.052	0.969a
P3a latency	313.44 ± 34.75	319.15 ± 37.63	0.917	0.361a
P3b latency	379.68 ± 14.90	357.78 ± 15.78	−6.966***	< 0.001b
ERP amplitude (µV)
N2 amplitude	9.45 ± 3.24	11.21 ± 3.38	−3.154**	0.002a
P3a amplitude	4.48 ± 0.96	4.67 ± 1.00	−1.129	0.261a
P3b amplitude	4.33 ± 0.83	4.52 ± 0.94	1.195	0.235a

SI Suicidal Ideation, NSI No-Suicide Ideation, FT3 Free Triiodothyronine, FT4 Free Thyroxine, TSH Thyroid Stimulating Hormone, COR Cortisol, TC Total Cholesterol, TG Triglycerides, LDL Low-Density Lipoprotein, HDL High-Density Lipoprotein, ERP Event-Related Potential

Values are presented as mean ± standard deviation

^aIndependent samples t-test

^bMann-Whitney U-test

^*P < 0.05, ^**P < 0.01, ^***P < 0.001

Predictors of SI in adolescents with first-episode depressive disorder

The logistic regression model demonstrated a good fit, supported by a non-significant Hosmer–Lemeshow test result (P > 0.1) and a Nagelkerke R² value of 0.729. The results of the multiple logistic regression analysis are detailed in Table 3. Ten variables were significant predictors of SI: gender, age, HAMA score, HAMD-24 score, positive coping score, negative coping score, TC level, P3a latency, P3b latency, and N2 amplitude. Risk factors for SI included: female gender (OR = 1.436, 95% CI: 1.128–1.743), older age (OR = 1.152, 95% CI: 1.059–1.245), higher HAMA score (OR = 1.872, 95% CI: 1.146–3.057), higher HAMD-24 score (OR = 1.215, 95% CI: 1.012–1.419), elevated TC (OR = 2.474, 95% CI: 1.369–4.471), higher negative coping score (OR = 1.877, 95% CI: 1.141–3.089), prolonged P3a latency (OR = 1.195, 95% CI: 1.034–1.382), and prolonged P3b latency (OR = 1.099, 95% CI: 1.063–1.137). Protective factors for SI included: higher positive coping score (OR = 0.764, 95% CI: 0.603–0.967) and greater N2 amplitude (OR = 0.796, 95% CI: 0.693–0.915).

Table 3.

The multivariate analysis of SI in 150 adolescents with first-episode depressive disorder

Variables	SI vs. NSI
	P	OR	95%CI
Gender (ref: male)	0.043	1.436	1.128–1.743
Residence (ref: rural)	0.832	0.841	0.423–1.672
Age	0.027	1.152	1.059–1.245
Age of onset	0.537	1.092	0.825–1.445
Disease duration	0.534	0.916	0.692–1.211
Educational attainment	0.253	1.428	0.776–2.630
HAMA	0.012	1.872	1.146–3.057
HAMD-24	0.015	1.215	1.012–1.419
Positive Coping Style	0.025	0.764	0.603–0.967
Negative Coping Style	0.013	1.877	1.141–3.089
FT3	0.411	2.588	0.331–4.843
FT4	0.521	0.818	0.444–1.510
TSH	0.442	0.539	0.112–2.602
COR	0.644	0.997	0.986–1.009
TC	0.003	2.474	1.369–4.471
TG	0.065	2.347	0.949–5.803
LDL	0.150	1.697	0.826–3.489
HDL	0.278	0.322	0.042–2.498
ERP latency
N2 latency	0.059	0.717	0.508–1.012
P3a latency	0.016	1.195	1.034–1.382
P3b latency	< 0.001	1.099	1.063–1.137
ERP amplitude
N2 amplitude	< 0.001	0.796	0.693–0.915
P3a amplitude	0.067	1.116	0.740–1.683
P3b amplitude	0.418	0.802	0.470–1.368

SI Suicidal Ideation, NSI No-Suicide Ideation, HAMA Hamilton Anxiety Scale, HAMD-24 Hamilton Depression Scale 24-item, FT3 Free Triiodothyronine, FT4 Free Thyroxine, TSH Thyroid Stimulating Hormone, COR Cortisol, TC Total Cholesterol, TG Triglycerides, LDL Low-Density Lipoprotein, HDL High-Density Lipoprotein, ERP Event-Related Potential

Discussion

Our study found that among adolescents with first-episode depressive disorder, female gender and older age are associated with increased SI. Several physiological and psychological factors were also associated with SI. Anxiety symptoms, depressive symptoms, negative coping style, TC levels, and P3a and P3b latencies were positively correlated with SI. Conversely, positive coping style and N2 amplitude were negatively correlated with SI. These psychological, physiological, and sociological markers collectively provide a multidimensional framework for understanding SI risk. Their integration enables a more holistic approach to early detection and personalized intervention strategies.

Globally, the incidence of suicide among adolescents continues to rise and is influenced by multiple risk factors [49]. In our study, the prevalence of SI among adolescents with first-episode depressive disorder was as high as 64.00%, significantly exceeding rates observed in the community adolescent group and the general population [50]. Consistent with previous studies on young individuals with depressive disorder, older age emerged as a risk factor for SI [51]. The heightened SI observed in late adolescence may be partly attributed to increased abstract reasoning ability during this period. Conceptualizing death abstractly may prompt individuals to be more capable of thinking about, planning suicide [52–54]. Moreover, older adolescents may experience greater autonomy and reduced parental supervision and discipline [55]. As they grow older, adolescents become more susceptible to peer influence in delinquent or risky behaviors [56]. Risk factors such as low-quality friendships and exposure to peer suicide may be more harmful to late adolescents, exacerbating the likelihood that they will escape their negative state through suicide [57].

Coping style, a modifiable psychological factor, serves as a key mechanism influencing suicidal ideation SI. Common coping styles used by adolescents include seeking social support, emotional support, distraction, self-blame, alcohol and tobacco use, and aggressive behaviors [58]. Poor emotion regulation, coupled with an inability to identify adaptive coping strategies, can perpetuate self-harm [59]. As also confirmed in our study that negative coping style is a risk factor for SI. Therefore, conceptualizing SI as a consequence of insufficient coping resources during emotion regulation and problem-solving, and incorporating coping skills training into suicide prevention strategies, may be effective in reducing SI in adolescents with depressive disorder.

Emotional disorders and lipid abnormalities may jointly increase suicide risk through neurobiological mechanisms. Anxiety and depressive symptoms can increase peripheral TC levels [60]. Moreover, these symptoms impair emotional regulation capacity by damaging the function of the prefrontal-limbic circuit, thereby exacerbating the severity of SI [61, 62]. Elevated peripheral TC levels may interfere with serotonin receptor and transporter function by altering cell membrane fluidity and causing vascular damage; these changes collectively lead to worsened negative emotions and increased suicide risk [63]. Emotional disorders and lipid abnormalities may form a mutually reinforcing vicious cycle, jointly depleting an individual’s psychological coping resources and ultimately increasing the risk of suicide in adolescents [64, 65].

Prolonged latency of P3a and P3b reflects cognitive deficits associated with SI. Stimulus signals in the frontal cortex produce P3a potentials related to attentional capture, and subsequently, stimulus signals are transmitted to the temporal and parietal lobes and produce P3b related to working memory, response inhibition, and executive function [40, 66]. In the context of a suicidal crisis, attentional capture and executive functioning are key cognitive processes that regulate emotion and manage SI [67, 68]. The coordination of the two helps individuals to flexibly use their attentional resources in the face of ambiguous or complex environments and to respond rationally by selectively attending to information and inhibiting irrelevant stimuli [69]. P3a abnormality may reflect impaired attentional control, where individuals have difficulty shifting attentional resources from negative thoughts to positive coping in the face of adverse stimuli such as negative emotions [70], while prolonged P3b latency may reflect cognitive rigidity in the executive domain, and irrational judgment of and reaction to internal and external environmental stressors [41]. It is speculated that individuals with prolonged latency of P3a and P3b, when faced with stressful events, are unable to rationally assess their situation, find it challenging to divert negative attention away from their predicament, and struggle to inhibit impulsive behaviors and ideation by intense emotions.

Decreased N2 amplitude may serve as an electrophysiological marker of SI in adolescents. A laser-evoked potential study showed that N2 amplitude was decreased in adolescents with NSSI compared to a healthy group [71]. N2 represents a cognitive process related to impulse detection, monitoring, and control [66]. Decreased N2 amplitude may indicate deficits in impulse regulation [39]. On the one hand, strong negative emotions can shape perceptions of self and future [72], and impulsive individuals with poorly constrained motivation and perceptions are more likely to be paralyzed by negative emotions and experience SI [73]. On the other hand, impulsive individuals tend to act emotionally and impulsively when experiencing negative emotions. Increased negative emotions exacerbate the urgency to avoid negative perceptions, leading to impaired impulse control [74]. Individuals who disregard the consequences of their actions may sacrifice long-term benefits (such as NSSI, SI, SB) in exchange for relief.

The multidimensional risk factors identified in this study have reference significance for addressing adolescent SI. While specific demographic factors such as female gender and older age are associated with increased risk, it is essential to emphasize that SI affects adolescents across all genders and age groups. Clinical practice and public health efforts must therefore adopt inclusive screening and intervention strategies, ensuring that support is accessible to all adolescents experiencing mental health challenges. The comprehensive intervention strategy must include psychosocial support, physiological monitoring, and neurocognitive training. Psychosocial support should focus on modifying maladaptive coping styles through targeted skills training, such as cognitive restructuring, problem-solving, and strengthening help-seeking behaviors. Clinical monitoring of blood lipids may help identify adolescents at heightened physiological risk. Prolonged P3 latency and reduced N2 amplitude highlight the potential of targeted cognitive training. Attention Bias Modification Training can help individuals with SI flexibly shift attention away from negative stimuli and reduce rumination [75]. Working Memory Training (WMT) has been shown to enhance executive function components, including reasoning ability and task-switching [76]. WMT may assist individuals with SI in maintaining goal-relevant information and suppressing irrelevant distractions more effectively when facing stress or negative emotions, thereby enhancing cognitive control.

Limitations

There are several limitations to this study. Firstly, a cross-sectional design limits causal inference due to the simultaneous measurement of variables. Secondly, data collection took place during the COVID-19 pandemic without accounting for pandemic-related factors. Additionally, distal risk factors for SI, such as genetic factors and family history, could not be adequately considered. Finally, assessing SI in adolescents solely through a single question in the HAMD-24 limits the accuracy of the results. The use of a validated, multidimensional SI assessment scale could enhance the precision and reliability of the assessment.

Conclusion

The diagnosis and treatment of SI in adolescents with depressive disorder represents a significant global public health challenge. SI among Chinese adolescents with first-episode depressive disorder is associated with factors such as gender, age, emotional state, coping style, blood lipid levels, and electrophysiological activity of the brain. The observed associations across physiological, psychological, and sociological domains suggest that a multidimensional framework could contribute to a more comprehensive understanding of SI in this population. Considering a range of risk and protective factors might inform the development of support frameworks for adolescents facing SI.

Abbreviations

SI

Suicidal Ideation

MDD

Major Depressive Disorder

SB

Suicidal Behavior

TSH

Thyroid Stimulating Hormone

TC

Total Cholesterol

LDL

Low-Density Lipoprotein

TG

Triglycerides

EEG

Electroencephalography

ERP

Event-Related Potential

NSSI

Non-Suicidal Self-Injury

ICD-10

International Statistical Classification of Diseases and Related Health Problems 10th Revision

HAMD-24

Hamilton Depression Scale 24-Item

HAMD-SI

Hamilton Depression Scale Suicide-Item

HAMA

Hamilton Anxiety Scale

SCSQ

Simplified Coping Style Questionnaire

FT3

Free Triiodothyronine

FT4

Free Thyroxine

HDL

High-Density Lipoprotein

COR

Cortisol

NSI

No-Suicide Ideation

WMT

Working Memory Training

Authors’ contributions

J Gao and W Li designed this study. Y Zhang and X Suo performed data analysis and wrote and revised this manuscript. J Xu, X Wang, W Xu, L Pan, and J Wang collected and organized the dataset. W Li and J Gao did the English revision. The final manuscript had been verified and approved by all authors.

Funding

This study was supported by the Qingdao Science and Technology Benefit People Program (Grant number: 22-3-7-smjk-19-nsh) and the Natural Science Foundation of Shandong Province (Grant number: ZR2021MH286).

Data availability

We do not have any research data outside the submitted manuscript file. The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study was conducted by the Declaration of Helsinki and was approved by the Ethics Committee of Shandong Daizhuang Hospital (Approval No. 202201KS-1). Participants were completely voluntary, anonymous, and based on written informed consent from the participants and their guardians, with the right to withdraw from the study at any time. To ensure participants’ privacy, all data collected was de-identified, and no personally identifiable information was collected or stored. Individual privacy was properly protected during survey administration and data processing.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.