James Donaldson on Mental Health – Association of unhealthy lifestyles and suicidal behaviors among Chinese middle school students: a latent class analysis

• Chunyan Wang, 
• Qiaoyue Wei, 
• Wenwen Yin, 
• Shuibo Pan, 
• Chenyangzi Dai, 
• Linhua Zhou, 
• Xiaoyu Jiang, 
• Hong Cai & 
• Junduan Wu 

Abstract

Background

The unhealthy lifestyles not only affect physical development of adolescents but are also closely related to various psychological problems. This article will examine potential classes of middle school students’ lifestyles and explore the associations between lifestyles and suicidal behaviors.

Methods

Data were collected from a cross-sectional survey of middle schools in five cities of Guangxi. A total of 5291 students completed the short questionnaire on suicidal behaviors and lifestyles. The latent class analysis was used to identify the unrecorded patterns of lifestyles. Multifactor logistic regression was used to explore the associations between health lifestyles and suicidal behaviors.

Results

The three classes lifestyles were identified: Class1(8.0%) – high risk lifestyle group, Class2(18.6%)—high lack of exercise group, and Class3(73.5%) – low risk lifestyle group, respectively.

Middle school students among Class 1 had higher odds than Class 3 (OR?=?2.061, 95%CI?=?1.474?~?2.882?OR?=?3.323, 95%CI?=?2.185?~?5.052 for self-injurious thoughts; OR?=?2.008, 95%CI?=?1.438?~?2.804?OR?=?3.217, 95%CI?=?2.125?~?4.868 for suicidal ideations; OR?=?1.488, 95%CI?=?1.063?~?2.081?OR?=?2.718, 95%CI?=?1.744?~?4.236 for self-injurious behaviors) for experiencing self-injurious thoughts, suicidal ideation, and self-injurious behaviors, whether occurring once, five times, or more, respectively. Class 2 were only more likely to engage in self-injurious behavior at once compared to Class 3(OR?=?1.414, 95%CI?=?1.110?~?1.800).

Conclusions

This finding highlighted the importance of improving unhealthy lifestyles for preventing and intervening in suicidal behavior among middle school students. More actions for intervention in the lifestyles of students should be needed to ensure their physical and mental health.

Peer Review reports

Introduction

Suicide is a major public health problem worldwide [31]. So far, numerous studies, both national and international, have demonstrated the increasing prevalence of suicidal behaviors among adolescents, with serious physical and psychological consequences [28, 52]. Suicide is the second leading cause of death among adolescents between the ages of 10 and 19 in the US [11]. However, in China, suicide is the leading cause of death in the young cohort aged 15–34 years, according to the 2023 National Health Commission (NHC) Report.

The unhealthy lifestyles

Unhealthy lifestyles, such as insufficient physical activity, poor dietary habits, and sleep deprivation, are the most common risk factors affecting the development and health of children and adolescents [54, 77]. Adolescence is a critical period for the establishment of lifestyle patterns [72]. The formative years of adolescence with lifestyle habits are often solidified, making it a critical window for intervention to prevent the adoption of unhealthy behaviors that may persist into adulthood. Studies have shown that unhealthy lifestyles in children and adolescents associated with physical health such as hypertension [69], diabetes [18], and dental caries [67], and mental health problems such as anxiety, depression, suicide and schizophrenia [37].

Given the multifaceted nature of unhealthy lifestyles and their far-reaching consequences, it is essential to explore lifestyle interventions for improving health outcomes across the physical and mental health spectrum. However, the literature related to this topic is not conclusive regarding the existence of specific underlying categories of unhealthy lifestyles among middle school students. This gap is significant, as understanding these categories can inform the development of targeted and effective interventions tailored to this age group population. Therefore, the aims of this study is to enrich this content.

Suicidal behaviors

Suicide is a continuum of behaviors, non-fatal suicidal behaviors include Suicidal ideation, Suicide attempt and suicide plan [64]. Suicidal ideation refers to the desire to die without taking actual action [10]. Suicidal ideation reflects individual suicide risk, and self-injurious thoughts and behaviors are important correlates of suicidal ideation. The prevalence of suicidal ideation, self-injurious thoughts and self-injurious behaviors among adolescents varies internationally: the lifetime prevalence of self-injury ranges from 21 to 33% in South Asian countries [29], and prevalence of self-injurious behaviors and suicidal ideation is 18.8% and 15.8% respectively in the United Kingdom [41]. In addition, cross-national studies have shown that the reported prevalence of suicidal ideation among adolescents was 21.1%, while self-injurious thoughts and behaviors were 9.55% and 2.83%, respectively [60].

Statistics shows that previous suicidal ideation, self-injurious thoughts and behaviors are the risk factor for suicide deaths [20, 65]. The results of suicide deaths highlight the importance of studying suicidal ideation, self-injurious thoughts and self-injurious behaviors. One of the aims of this article is to investigate the current prevalence of self-injurious thoughts, suicidal ideation and self-injurious behaviors among Chinese middle and high school students. Previous studies have explored the psychosocial, environmental, psychological and biological risk factors for adolescent suicide [12, 63]. Relatively few studies, however, have systematically explored patterns of lifestyles that precede or coincide with suicidal behaviors.

The lifestyles and suicidal behaviors

Low levels of physical activity, insufficient sleep, high consumption of sugary drink, and low intake of fruits and vegetables are common unhealthy lifestyle behaviors among adolescents, which may increase the risk of suicidal behaviors in this population. Specifically, previous research has demonstrated that increased physical activity is linked to reduced suicidal ideation and self-harm [24, 58]. Furthermore, the lack of sleep was a risk factor for self-harm [35]. High levels of soft drink consumption have been associated with suicidal behaviors [38, 46, 59]. Several studies have shown the correlation between mental health and the consumption of fruits and vegetables [22]. Individuals with a history of attempted suicide have been found to consume significantly less fruits and vegetables [45]. Additionally, some studies have investigated the collective influence of one or two specific types of health behaviors on suicide-related behaviors. For instance, a national sample of adolescents found that physical activity, sleep, and the school environment significantly predict suicidal ideation [61], adolescents with a lifestyle risk score of 5 or 6 are 3.36 times more likely to have suicidal ideation than those with a lifestyle risk score of 0 [37]. Unhealthy lifestyle practices may contribute to increased inflammatory responses in the body [25] which in turn can impact suicidal behaviors [66].

Research on the effects of lifestyles on suicide-related behaviors among Chinese middle students is limited, and previous research has primarily focused on individual health behaviors, overlooking the combined effects of multiple lifestyle factors on different suicidal behaviors. Our study will employ latent class analysis to identify student subgroups with similar lifestyle characteristics, thereby exploring the complex associations between lifestyle patterns and suicidal behaviors among Chinese middle school students. This approach contributes to a deeper understanding of the intricate relationships in this area.

The present study

The study focuses on three key domains of health behaviors related to daily activities that play a crucial role in suicidal behaviors: dietary behaviors (sugary drinks and fruit and vegetable intake), physical activity, and sleep. The purpose of this study is threefold, Firstly, to evaluate the prevalence of self-injurious thoughts, suicidal ideation and self-injurious behaviors among Chinese middle school students; secondly, to identify different lifestyle latent classes and to provide the socio-demographic and psychosexual-related characteristics of the latent classes; and finally, to explore the relationship between lifestyle latent classes and self-injurious thoughts, suicidal ideation and self-injurious behaviors among Chinese middle school students. It is worth mentioning that in our study, the relationship between lifestyle and suicidal behaviors is conceptualized through Latent Class Analysis (LCA). This method enables us to identify student groups with distinct lifestyle patterns and examine the association between these patterns and suicidal behaviors.

Methods

Participants

In 2022, a cross-sectional study was conducted using a multi-stage stratified random sampling method [1, 68]. Firstly, five cities (Nanning, Liuzhou, Beihai, Guigang, Hechi) were selected based on their level of economic development in Guangxi. Secondly, 1–2 urban areas or counties were randomly selected from each city. Thirdly, 1–2 middle schools were randomly selected from each urban/county. Finally, two classes from each grade of each school were selected for a questionnaire survey. Prior to the survey, both students and their legal guardians were fully informed of all relevant investigation content and procedures, and provided their written informed consent. Excluding students who refused to participate in this study and those who were absent that day, a total of 5331 participants aged 14 to 19 were initially included in the study. However, due to incomplete and logically flawed survey responses, 213 questionnaires were excluded from the analysis. The final sample size consisted of 5291 student questionnaires that met the study’s inclusion criteria. The response rate was high at 99.25% (5291/5331). This study was approved by the Institutional Ethical Committee of Guangxi Medical University.

Measurements

Participants’ general information

The study collected basic demographic information from participants, including age, gender, race/ethnicity, place of residence, only child status, caregivers, gender identity, sexual orientation, parental affection, and parental absence of both father and mother. Furthermore, this study defines left-behind adolescents as individuals who have been left behind at their original residence while one or both parents work outside the family home for more than six months [19].

The lifestyles indicators

This study measured middle school students’diet, physical activity and sleep using four items. These three domains of health behaviors were actions and indicators covered in the Healthy China Action (2019–2030) (Healthy China Action (2019-2030), [32]). All the cutoff scores were based on recommendations from the Chinese Dietary Guidelines and significant thresholds used in previous studies on lifestyles and suicidal behaviors [15].

Frequency of exercise were assessed with a standard question’On average how frequently do you exercise each week?’, and included options categorized as a five- point scale:”1? means”hardly ever exercise”,”2? means”1 time”,”3? means”2–3 times”,”4? means”4–5 times”, and”5? means”more than 5 times”. Participants were classified as in health status if their frequency of exercise scores were ??3 points; Frequency of consumptions of fruits and vegetables was assessed with a standard question’On average, how frequently do you consume fruits and vegetables each week?’, and included options categorized as a five- point scale:”1? means”eating 0 time”,”2? means”1–2 times”,”3? means”3–4 times”,”4? means”5–6 times and”5? means”more than 7 times”. The participants were classified as healthy if their frequency of fruit and vegetables consumption scores equaled five points in our study; Frequency of consumptions of sugary drinks was assessed with a standard question’On average, how frequently do you consume sugary drinks per week? Such as Coke, Sprite, orange juice, milk tea, etc.’, and included options categorized as a five- point scale:”1? means”never”,”2? means”1–2 times”,”3? means”3–4 times”,”4? means”5–6 times and”5? means”more than 7 times”. Participants were classified as healthy if their frequency of sugary drinks consumption scores ??2 points in our study; Frequency of sleep deprivation was assessed with a standard question:’Have you experienced sleep deprivation in the past month?’, and included options categorized as a five- point scale:”1? means”never”,”2? means”1 time per week”,”3? means”1–2 times per week”,”4? means”3–5 times per week “and”5? means”5–7 times per week”. Participants were classified as Sleep deprivation if their scores ??3 points in our study. The four lifestyle-related survey questions we utilized have been extensively used in China [27, 44, 47, 79].

Suicidal behaviors

Suicidal behaviors were assessed with the following three questions: (1) “Have you ever thought about deliberately injuring yourself when you don’t want to die? Like slitting your wrists.” (Self-injurious thoughts), (2) “Have you ever considered suicide?” (suicidal ideation), and (3) “Have you ever injured yourself?” (Self-injury behaviors). The survey questions we utilized have been widely applied among Chinese middle school students [73, 78]. In order to reduce variability in suicidal behaviors frequency estimates, number of suicidal behaviors were categorized [9, 74] by the following frequencies: 0 time, 1 time, 2–4 times, and 5 or more times.

Covariates

Covariates included demographic and psychosexual related variables. Demographic information regarding age (continuous, 14 ~?19 years), sex (boy/girl), nationality (Han, Zhuang, other), place of residence (city?county?rural),only child status (yes/no), caregivers (parents/grandparents/other), parental affection(good/bad) and staying behind(yes/no) were collected during the self-report assessment. Previous studies have suggested that gender, age, ethnicity, place of residence, only child, caregiver, parental affection and leaving behind are important factors in adolescent suicide [6, 21, 33, 36]. Gender minority students and sexual minorities had significantly higher rates of suicide ideation and suicide attempts [34]. Gender identity and sexual orientation variables were included to control for their effects. The survey asked students to indicate their gender identity and sexual orientation. The questions were phrased as follows:’Which gender do you identify with?’and’Which of the following options best describes your sexual orientation? (choose only one)”. Response options were dichotomized into “male, female, none of the above”, and “heterosexual, homosexual, bisexual, and uncertain”.

Statistical analysis

Descriptive statistics

Statistical assumptions were checked for each variable prior to conducting the analyses. The level of significance was set at p?< 0.05, and all assumptions were met through chi-square tests, t-tests, and analysis of variance (ANOVA). Descriptive statistics for the demographic characteristics and psychosexual related variables of the study sample and comparisons across groups were analyzed using SPSS software version 23.0.

The latent class analysis (LCA)

The latent class analysis(LCA) was performed using Mplus version 8.3 to identify and describe the classes of students’ health lifestyles. In order to determine the exact number of latent classes, an exploratory approach was used [75]. Starting from a two-class model, the analysis was carried out several times in a row, increasing this number of subgroups turn by turn. Then, statistical model adjustment indices, such as the Bayesian information criterion (BIC), the adjusted BIC, the Akaike information criterion (AIC), LoMendell-Rubin likelihood ratio test (LMR), and Bootstrapped likelihood ratio test (BLRT) were used, coupled to the entropy values [56, 70] to identify the final number of classes. Indeed, the selected classes had to have more than 5% observations to provide a representative profile of a population [75].

Bivariate analyses and multivariate logistic regression

After determining the appropriate number of latent classes, we conducted a series of cross-tabulations and bivariate analyses to explore the distribution of demographic and psychosexual-related characteristics across different latent classes. Additionally, we used univariate analysis to explore the influencing factors of suicide related behaviors. Finally, we adjusted for demographic and psychosexual characteristics based on the results of the univariate analysis. Multiple logistic regression analysis was used to investigate the influence of lifestyle latent classes on suicide-related behaviors. A statistically significant difference was defined as a two-sided p-value of less than 0.05.

Results

The prevalence of suicidal behaviors and general demographic data

A total of 5291 participants were included in this study, and their gender distribution was relatively average (47.3% Boys and 52.7% Girls). The respondents were predominantly Han Chinese students (54.8%), followed by Zhuang (29.1%), and other ethnic students (16.1%). The mean age of the students was 15.65 (Standard Deviation, SD =?1.353). The prevalence of self-injurious thoughts, suicidal ideation, and self-injurious behaviors among middle school students is 14.6%, 27.7%, and 17.0%, respectively. For sexual orientation, those who had self-injurious thoughts and suicidal ideation had the highest rate of homosexual students (42.6%; 60.3%)), however, those who had self-injurious behaviors had the highest rate of bisexual students (42.5%). Table 1 summarizes these results.Table 1 Demographic characteristics of suicidal behaviors

Full size table

Modeling of latent classes for adolescent healthy lifestyles

This study identified three-class model due to its entropy value (0.779) exceeding the standard for good class separation (i.e. entropy =?0.60; [4]) and having the smallest AIC, BIC, adjusted BIC and appropriate LMR and BLRT (P < 0.0001). Table 2 describes these indicators.Table 2 Summary of latent class model identification and fit statistics

Full size table

Figure 1 shows the three-class model and item probabilities for the 3 lifestyles indicators for each class. Class1, defined as the”high risk lifestyle”(8.0%), included adolescents reporting the medium lack of exercise; lower levels of fruit and vegetable intake, higher intake of sugary drinks and serious sleep deprivation. Class2, defined as the”high lack of exercise”(18.6%), included adolescents reporting the high lack of exercise, lower levels of fruit and vegetable intake, lower intake of sugary drinks, and medium sleep deprivation. Class3, defined as the”low risk lifestyle”(73.5%), included adolescents reporting the high levels of exercise, high levels intake of fruit and vegetable, lower intake of sugary drinks and low levels of sleep deprivation.

Distribution of suicidal behaviors and covariates by lifestyle classes

Table 3 shows significant differences in the distribution of suicidal behaviors, demographic characteristics, and psychosexual-related characteristics among the 3-class model.Table 3 Self-injurious thoughts, Suicidal ideations, Self-injury behaviors and demographic characteristics by latent classes

Full size table

Regarding suicidal behaviors, class 1 (high-risk lifestyle) had higher percentages of self-injurious thoughts (11.9%, 5.5%, 9.0%), suicidal ideation (15.4%, 12.4%, 17.1%), and self-injurious behaviors (11.4%, 7.1%, 7.4%) compared to the other classes (P?< 0.001). In contrast, class 3 (low-risk lifestyle) had the lowest percentages (P?< 0.001). Significant differences were found among the three models in terms of age, gender, ethnicity, place of residence, only child status, parental affection, gender identity and sexual orientation (P?< 0.05).

The lifestyles and suicidal behaviors

Single?factor analysis of suicidal behaviors

We used the chi-square test and ANOVA to determine differences about three different suicidal behaviors among variables. Statistical significance was found in self-injurious thoughts among adolescents of different age, gender, place of residence, the only child, caregivers, gender identity, sexual orientation, parental affection, staying behind (P?< 0.01). Then, there was statistical significance in suicidal ideation among adolescents of different gender, place of residence, the only child, gender identity, sexual orientation, parental affection and staying behind (P?< 0.01). And Statistical significance was found in self-injury behaviors among adolescents of different age, gender, place of residence, gender identity, sexual orientation, parental affection and staying behind (P?< 0.01). An additional table file shows them in more detail [see Additional file 1].

Multivariate logistic regression analysis of suicidal behaviors

Only variables with statistically significant results were included in the multivariate logistic regression analyses for all suicide-related behaviors, with no suicide-related behaviors as the reference level.

For self-injurious thoughts, compared to class 3 (low-risk lifestyle), class 1 (high-risk lifestyle) was approximately 2.061 times more likely to have self-injurious thoughts 1 time (95% CI =?1.474 ~?2.882), and 3.323 times more likely to have self-injurious thoughts 5 or more times (95%CI =?2.185 ~?5.052).

However, there was no statistical significance for class 2 (high lack of exercise group).

Meanwhile, our study found that younger students had a lower risk of reporting self-injurious thoughts 1 and 2?~ 4 times (OR =?0.918, 95%CI =?0.843 ~?0.999; OR =?0.849, 95%CI =?0.762 ~?0.946).

Regardless of the number of suicidal thoughts, boys had a lower risk of self-injurious thoughts compared to girls (OR =?0.569, 95%CI =?0.451 ~?0.718; OR =?0.410, 95%CI =?0.302 ~?0.557; OR =?0.326, 95%CI =?0.219 ~?0.485).

Regardless of the number of suicidal thoughts, students with affectionate parents had a lower risk of reporting self-injurious thoughts (OR =?0.536, 95%CI =?0.421 ~?0.683; OR =?0.448, 95%CI =?0.328 ~?0.613; OR =?0.352, 95%CI =?0.233 ~?0.532).

Compared to heterosexual adolescents, homosexual and bisexual students had a higher risk of reporting self-injurious thoughts (OR =?2.706 ~?5.096, 95%CI =?1.194 ~?11.542; OR =?1.780 ~?3.840, 95%CI =?1.180 ~?6.087). Table 4 summarizes these results.

Table 4 Model estimates predicting self-injurious thoughts

Full size table

For suicidal thoughts, compared to class 3 (low-risk lifestyle group), class1 (high-risk lifestyle group) was 2.008 times more likely to have one time self-injurious ideation (95% CI =?1.438 ~?2.804), and 3.217 times more likely to have 5 or more times self-injurious ideation (95% CI =?2.125 ~?4.868).

However, there was no statistical significance for class 2 (high lack of exercise group). Meanwhile, our study found that regardless of the number of suicidal thoughts, boys had a lower risk of suicidal ideation compared to girls (OR =?0.570, 95%CI =?0.451 ~?0.719; OR =?0.405, 95%CI =?0.298 ~?0.550; OR =?0.334, 95%CI =?0.225 ~?0.495).

Compared to cisgender adolescents, non-binary adolescents had a higher risk of reporting 2?~ 4 or more times suicidal ideation (OR =?2.120, 95%CI =?1.194 ~?3.766; OR =?2.577, 95%CI =?1.431 ~?4.640).

Compared to heterosexual adolescents, homosexual and bisexual adolescents had a higher risk of reporting suicidal ideation (OR =?3.003 ~?5.356, 95%CI =?1.332 ~?12.135; OR =?1.829 ~?3.910, 95%CI =?1.213 ~?6.183).

Regardless of the frequency of suicidal ideation, adolescents with affectionate parents had a lower risk of reporting suicidal ideation(OR =?0.536, 95%CI =?0.421 ~?0.682; OR =?0.448, 95%CI =?0.328 ~?0.612; OR =?0.351, 95%CI =?0.233 ~?0.530). Table 5 summarizes these results.

Table 5 Model estimates predicting suicidal ideations

Full size table

For self-injurious behaviors, compared to class3 (low-risk lifestyle group), class1 (high-risk lifestyle group) was

1.488 times more likely to have one time self-injurious behavior (95% CI =?1.063 ~?2.081),

and 2.718 times more likely to have self-injurious behaviors 5 or more times (95% CI =?1.744 ~?4.236),

and class2 (high lack of exercise group) was 1.414 times more likely to have one time self-injurious behavior (95% CI =?1.110 ~?1.800).

Meanwhile, our study found that younger adolescents had a lower risk of reporting self-injurious behaviors 5 or more times (OR =?0.867, 95%CI =?0.759 ~?0.990). Regardless of the frequency of self-injurious behaviors, boys had a lower risk of self-injurious behaviors compared to girls (OR =?0.554, 95%CI =?0.447 ~?0.685; OR =?0.460, 95%CI =?0.350 ~?0.603; OR =?0.294, 95%CI =?0.195 ~?0.444).

Compared to cisgender adolescents, transgender adolescents had a higher risk of reporting 1 or 2?~ 4 times self-injurious behaviors (OR =?1.814, 95%CI =?1.247 ~?2.637; OR =?1.759, 95%CI =?1.145 ~?2.703).

Compared to heterosexual adolescents, homosexual and bisexual adolescents had a higher risk of reporting 2?~ 4 or more times self-injurious behaviors (OR =?2.993 ~?3.405, 95%CI =?1.184 ~?7.566; OR =?3.947 ~?4.516, 95%CI =?2.739 ~?7.073).

In addition, bisexual adolescents had 1.527 times more likely to have self-injurious behavior 1 time (95%CI =?1.017 ~?2.293). Regardless of the frequency of self-injurious behaviors, adolescents with affectionate parents had a lower risk of reporting self-injurious behaviors (OR =?0.491, 95%CI =?0.392 ~?0.614; OR =?0.429, 95%CI =?0.323 ~?0.569; OR =?0.390, 95%CI =?0.262 ~?0.580). Table 6 summarizes these results.

Table 6 Model estimates predicting Self-injurious behaviors

Full size table

#James Donaldson notes:
Welcome to the “next chapter” of my life… being a voice and an advocate for #mentalhealthawarenessandsuicideprevention, especially pertaining to our younger generation of students and student-athletes.
Getting men to speak up and reach out for help and assistance is one of my passions. Us men need to not suffer in silence or drown our sorrows in alcohol, hang out at bars and strip joints, or get involved with drug use.
Having gone through a recent bout of #depression and #suicidalthoughts myself, I realize now, that I can make a huge difference in the lives of so many by sharing my story, and by sharing various resources I come across as I work in this space.  #http://bit.ly/JamesMentalHealthArticle
Find out more about the work I do on my 501c3 non-profit foundation
website www.yourgiftoflife.org Order your copy of James Donaldson’s latest book,
#CelebratingYourGiftofLife: From The Verge of Suicide to a Life of Purpose and Joy

Click Here For More Information About James Donaldson

Discussion

The study used LCA to identify three distinct lifestyles among Chinese middle school students, each characterized by varying levels of engagement in health-promoting and health-risk behaviors. Individuals with moderate-risk and high-risk lifestyles (i.e., Class 2 and Class 1) exhibited significantly higher risks of suicide related behaviors compared to those with the low-risk lifestyle (i.e., Class 3). Differences in demographic characteristics and psychosexual related variables were observed across lifestyle latent classes, supporting The Healthy Lifestyles Theoretical Model. To our knowledge, this study is the first in Guangxi Zhuang Autonomous Region to investigate the relationship between lifestyles and suicidal behaviors among middle school students with representative data from the entire province.

The latent class of lifestyles

Among Chinese middle school students, three different lifestyles were identified using LCA. This study suggests that middle school students between the ages of 14 and 19 have distinct categorical characteristics in their lifestyles. These characteristics occur in clusters and in varying proportions. Issues with exercise levels, eating habits, and sleep patterns among middle school students are prevalent, and sleep deprivation is a particularly significant concern. In the context of Chinese middle school students, our study, aligned with previous research, identifies regular physical activity (frequency of exercise ??2–3 times/week, a balanced diet (consumption of fruits and vegetables ??7 times/week, consumption of sugary drinks ??1–2 times/week), and adequate sleep (no sleep deprivation) as behaviors that are widely accepted and encouraged within Chinese society for the maintenance of good health and well-being [14,15,16]. Foreign research has classified healthy lifestyles into three latent classes [39]. Similarly, a recent study has classified the lifestyle of Chinese middle school students into three classes: Class1 (high risk behavior), Class2 (low lack of exercise/high intentional injury), and Class3 (low risk behavior) [51], this is in line with the current study. Timely and targeted interventions by relevant departments are necessary to reduce the prevalence of unhealthy lifestyles among middle school students.

High risk of lifestyle and suicidal behaviors

Our findings indicate that students with a medium exercise level, high unhealthy diet, and serious sleep deprivation (Class 1) have the highest risk of suicide-related behaviors. Previous reports have shown that unhealthy lifestyle behaviors and mental health problems are closely related [3, 49]. According to the suicide avoidance theory, feelings of unhappiness and disgust arise when an individual perceives themselves as failing to meet significant standards, which can predispose them to suicidal behavior [5]. Here, the significant standards emerge, evolve, and are solidified through the processes of observation, interaction, and communication within a group. In other words, collective perceptions regarding the frequency and acceptability of certain behaviors can crystallize, leading to the establishment of normative patterns of conduct [48]. Among middle school students, the failure to adhere to perceived health standards, including a balanced diet and sufficient sleep, can lead to feelings of unhappiness and disgust. Prolonged and intensified experiences of these negative emotions may escalate to the point where they precipitate suicidal behaviors.

Other possible explanations for the high rate of suicidal behaviors among Class1 students may include their specific lifestyle, which is characterized by severe sleep deprivation. Studies have shown that Class1 students have the highest probability of self-reported sleep deprivation, which is significantly associated with suicidal behaviors [23, 62]. In China, this is particularly the case. For example, studies have identified sleep problems as an independent influencing factor for suicidal behavior among Chinese middle school students [13], and middle school students who sleep less than 8 h per day have risks of suicidal ideation, suicide planning, and suicide attempts that are 1.152, 1.182, and 1.263 times higher, respectively, than those who sleep between 8 to 10 h [78]. Additionally, it is well known that sleep deprivation affects the activity of the hypothalamic–pituitary–adrenal (HPA) axis. Studies have shown that severe sleep deprivation may lead to HPA axis dysfunction, which could be associated with the suicidal process [7, 26, 53, 57]. However, the exact mechanisms of this association are not yet clear, and further research is needed to examine the link between high-risk unhealthy lifestyle and suicidal behaviors.

Medium lifestyles and suicidal behaviors

Compared to Class 3, our study found Class 2 is more likely to report one-time self-injurious behavior. A possible explanation is their specific health-promoting behaviors. Class2 students have the lowest odds of meeting the standard in terms of exercise levels. Consistent with Yuan et al.’s study on Chinese middle school students [78], research from other countries has also demonstrated a negative correlation between physical activity and self-injurious behavior, indicating that physical activity can effectively reduce the incidence of self-injurious behavior [2, 62].

Other potential explanations relate to the specific nutrients found in fruits and vegetables. Consistent with previous studies targeting Chinese middle school students, we define a normal intake of fruits and vegetables as at least once per day [80]. The high levels of consumption of fruits and vegetables may reduce the risk of self-injurious behaviors, it is possibly due to the high levels of bioactive compounds they contain, such as dietary fiber, vitamins, mineral antioxidants and flavonoids. These compounds can reduce inflammation and oxidative stress in the brain [8], leading to improved psychological well-being. However, further research is needed to fully understand the mechanisms behind this correlation.

Low-risk lifestyles and suicidal behaviors

In our study, Class 3 exhibited the healthiest lifestyle and the lowest risk of suicide-related behaviors. This finding is also consistent with previous reports that healthier lifestyles can lower the risk of suicide, participants who reported good sleep and high levels of physical activity were also among those who reported low self-injurious behavior [40].

In conclusion, our study indicates that adolescent suicide-related behaviors may be influenced by lifestyles with low levels of engagement in health-promoting behaviors and high levels of engagement in health risk behaviors. Therefore, we recommend that educational and health authorities implement mandatory health education programs to promote mental well-being among middle school students, particularly through systemic, school-based interventions. These interventions could include improving physical education programs, promoting physical activity during school hours, adjusting school schedules to ensure sufficient sleep, and fostering a supportive school environment. ?As part of creating a supportive school environment?, efforts should be made to enhance access to mental health services and develop school-wide initiatives that encourage balanced diets and stress management techniques. It is crucial to continuously monitor and evaluate these interventions to ensure their effectiveness in creating a healthy and supportive environment for students. Prioritizing these actions will help us to mitigate the risks associated with unhealthy lifestyle choices among students, thereby safeguarding their physical and mental health.

Demographic differences across classes

The demographic characteristics of the three healthy lifestyles indicate significant health disparities. Firstly, with regard to lesbian, gay, bisexual, and transgender (LGBT), we found that transgender and bisexual students were more likely to belong to the unhealthiest lifestyle (Class3), while homosexual adolescents were more likely to enter the moderate healthy lifestyle group (such as Class2). Previous studies have also found differences in diet, sleep and exercise levels among LGBT [42, 50]. However, other studies found that there was no significant difference between heterosexual and LGBT people in terms of exercise levels, dietary disorders, and alcohol misuse [17, 55]. Therefore, more researches are needed to supplement and expand the lifestyle of LGBT people.

Secondly, Consistent with previous literature [43, 76], the likelihood of ethnic minorities engaging in the most harmful lifestyles (e.g. class 1) is higher than that of Han Chinese. Therefore, in order to solve the unique health disparities faced by minority adolescents., it is necessary for us to emphasize the differences in culture and living habits.

Finally, Girls and older adolescents are more likely to enter the unhealthy lifestyle groups (e.g. Class1 and Class2), rather than the healthiest lifestyle group (e.g. Class3). The level of physical activity in girls was significantly lower than that in boys [71], which may mean that the lack of exercise level may be an important factor for girls to enter the unhealthy lifestyle group. Furthermore, Recent research findings support the influence that physical activity, dietary structure, and sleep tend to deteriorate with age [30]. It is important to note that these factors do not necessarily cause suicidal behavior, but they may increase the risk.

Strengths and limitations

Previous studies on healthy lifestyles have generally relied on cluster analyses or cumulative risk scores, but our study employed more innovative approach (i.e. LCA). It categorizes individuals based on their different patterns of response to the apparent variables, and is person-centered and takes into account individual variability. In addition, this study employs a combination of LCA and regression analyses to examine the latent categories of healthy lifestyles and their associations with suicidal behaviors. This provides a promising direction for researchers in designing suicide prevention and intervention plans, as well as a possible basis for clinical diagnosis, which has important practical significance.

Nevertheless, there are some limitations. Firstly, as the subjects were interviewed retrospectively by questionnaire, there is some recall bias; some items (e.g. sexual orientation, etc.) may be considered sensitive and therefore actual exposure may have been underestimated. Second, the cross-sectional design of this study precludes definitive conclusions regarding causality. Third, the study’s focus on middle school students in Guangxi Province, China, limits the generalizability of the findings to other regions or cultural contexts.

Conclusion

Our research reported that there are three potential classes of lifestyles among middle school students. High risk lifestyle is the risk factor of self-injurious thoughts, suicidal ideation and self-injurious behaviors. Additionally, high lack of exercise is the risk factor of self-injury behaviors.