Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

A comparative analysis of suicide attempts in left-behind children and non-left-behind children in rural China

  • Hongjuan Chang,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Qiuge Yan,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Lina Tang,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Juan Huang,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Yuqiao Ma,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Xiaozhou Ye,

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

  • Yizhen Yu

    Affiliation The Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, Wuhan, Peoples Republic of China

A comparative analysis of suicide attempts in left-behind children and non-left-behind children in rural China

  • Hongjuan Chang, 
  • Qiuge Yan, 
  • Lina Tang, 
  • Juan Huang, 
  • Yuqiao Ma, 
  • Xiaozhou Ye, 
  • Yizhen Yu


To estimate the prevalence of suicide attempts and explore the shared and unique factors influencing suicide risk in left-behind children (LBC) and non-left-behind children (NLBC) in rural China, this study collected data using a multi-stage cluster random sampling method from 13,952 children including 6,034 LBC and 7,918 NLBC. Sociodemographic characteristics, suicide attempts, neglect and physical abuse, negative life events, and loneliness were measured by self-reported questionnaires. Data were analyzed using logistic regression models. Gender and mother's education level were unique influential factors for NLBC while family structure type was a unique influential factor for LBC. The study provides two novel findings regarding NLBC specifically: 1. Children with optimal family socioeconomic status are more likely to report suicide attempts (odds ratio OR = 1)than are those in the general children population, OR 0.52 (95% CI: 0.39–0.70), and 2. Children with higher mother’s education level are subject to higher suicide rates in high school, OR 1.67 (95% CI: 1.13–2.46), and post-secondary education, OR 2.14 (95% CI: 1.37–3.37). The unique characteristics of LBC and NLBC in China suggest that investigating risk factors and determining the factors that might be targeted in intervention programs are urgently needed currently.


Suicide is a global health problem and a major public health concern. It is among the top causes of mortality worldwide, especially among adolescents in the West and developing countries [1]. Since suicide is a potentially preventable public health issue, it is important to examine its immediate precursors, especially suicide attempts (SAs), which refer to direct efforts to intentionally end one's own life, to aid in the development of future public health interventions [2,3].

Adolescents’ suicidal behavior has been reported to be associated with genetic, psychological, social, and familial factors with particular risks related to childhood adversities [4]. Many studies have pointed out a strong and graded association between exposure to adverse childhood experiences and SAs during adolescence and adulthood [46], and the most concerning adversities for suicidal behavior are abuse and neglect [7,8]; in developing countries, large numbers of adolescence have been exposed to these [9,10]. Evidence suggests that those who perceived their parents as almost always emotionally neglectful had an increased risk of psychiatric disorder including SAs [11]. Other major psychosocial factors related to suicidal behavior included the experience of negative events during the past 12 months and feelings of loneliness. Research demonstrated that attempters had experienced a greater number of negative life events prior to their attempts [12]. Furthermore, there was a significant dose—response relationship between negative life events experienced within the last year and increased risk of SAs [13]. In addition to negative events, it may be useful to consider the role of loneliness in predicting suicide risk in adolescents. Looking at correlates of suicide risk in existing studies, we found that children and adolescents who experienced chronically high levels of loneliness were more likely to report suicidality [14,15].

Left-behind children (LBC) are those rural children under 18 years of age who are left at home when one or both parents migrate to an urban area for work [16]. Since China has recently experienced unprecedented urbanization and changes in the structure of society, foreshadowing a population of rural-to-urban migrants and the offspring they leave behind [17], several studies have focused on the problems of LBC in this area [8,18]. The literature in general have shown that LBC had more psychological problems than NLBC, including inadequate family bonding, emotional vulnerability, and exposure to violence [8,19].

However, to date, it remains unclear whether the rate of SAs is higher in LBC than in NLBC in rural China among nationally representative samples. Nor has any research examined the specific difference of SA risk factors between LBC and NLBC within the current social context of China.


SPSS (version 17.0) was used for data analysis. Chi-squared tests were used to compare the difference of enumeration data variables across LBC and NLBC groups. Logistic regression analysis was performed to calculate the odds ratios (OR) and 95% confidence intervals (95% CI) for the factors related to SAs. All of the factors were chosen as the independent variables. All tests were two-tailed, and a P-value smaller than 0.05 was considered statistically significant.

Materials and methods

Participants and procedures

The data of the study were part of a nationwide study on mental health outcomes among adolescents in rural China, which was initiated in 2015. A multi-stage cluster random sampling method was adopted to collect data to represent all students from 7th to 12th grade in rural China. At the first stage of sampling, five districts were selected from the north, south, east, west, and middle part of China to represent the whole rural area of China. In the second stage, three counties or cities were chosen randomly in each province. In the last stage, schools were selected basing on their reported enrollment size. Excluding those who refused to participate in the study and who were absent from school, a total of 15,600 students were recruited in our study, and a consent letter was sent to their parents or guardians. Among the respondents, 1,648 were excluded due to incomplete questionnaires. Finally, 13,952 students were included in our analysis: 6,034 LBC and 7,918 NLBC with a mean age of 15.22 (SD = 1.81) years, ranging from 10 to 18 years. The actual response rate of the participants was 89.43% (13,952/15,600). Informed consent was obtained from the parents or guardians of each participant.

Data were collected by a group of trained postgraduate students, who explained the purpose and procedures of the study to participants. The students were instructed to place not their names but their student numbers on the questionnaires and to answer all the questions honestly. The students were also informed that their participation was voluntary and that the questionnaires did not represent a test as there were no correct or incorrect answers. We told each participant that they were to place the questionnaires in envelopes after completion and were not to hand them directly to school teachers or school personnel, and we promised them that the data would be used for scientific research only. The same written announcements were printed on the front of the questionnaires. The ethical protocol, including the questionnaires, was approved by the targeted schools and the Medical Ethics Committee of Tongji Medical College, Huazhong University of Science and Technology. Informed written consent was obtained from the parents or guardians of each participant.


Sociodemographic characteristics.

Information about gender, age, class, grade, school, and family structure type as well as education levels of parents and caregivers, family's socioeconomic status, parenting style, and family history of mental illness was collected.

Suicide attempts.

Information on SAs was collected by the question, “During the preceding 12 months, how many times did you actually attempt suicide?” For this item, responses fall into two categories: “never” and “more than once.” The test—retest reliability of the question over two weeks was 99.41% in the present study.

Neglect and physical abuse.

The prevalence of neglect and physical abuse was measured by a validated Chinese version of the Parents—Child Conflict Tactics Scale (CTSPC), which was developed to assess the subjective feelings of neglect and physical abuse among children or adolescents [20]. The scale consists of 17 items on four factors: neglect (e.g., ‘‘parents left you alone when you were in need of their company”), corporal punishment (e.g., ‘‘parents slapped you on the hand, arm, or leg”), physical maltreatment (e.g., ‘‘parents hit you with a fist or kicked you hard”), and severe physical maltreatment (e.g., ‘‘parents grabbed you around the neck and choked you”). The students in our study were asked how frequently they had encountered the listed behaviors in the past year using a three-point Likert scale covering none, once, and twice or more. Their responses were classified as “neglect or maltreatment of a particular type” if they had experienced one or more of the listed behaviors within the corresponding subscale. The neglect subscale consists of five items that range from 0 to 10. We defined scores of 0, 1–3, 4–6, and 7–10 as no, mild, moderate, and severe neglect, respectively. We defined the participants who did not experience the three subscales of physical abuse as having no physical abuse. Children who experienced corporal punishment only were defined as having mild physical abuse. Children who experienced physical maltreatment were defined as having moderate physical abuse, and children who experienced severe physical maltreatment were defined as having severe physical abuse. The internal consistency score of the whole scale was 0.86.

Negative life events.

The Adolescent Self-Rating Life Events Checklist (ASRLEC), a 5-point Likert scale, aims to assess whether certain life events occurred to the participant as well as the effects, if any, in the past 12 months [21]. The scale consists of 27 items on six factors: interpersonal relationship (e.g., “I argued with my classmates”), study pressure (e.g., “I failed an examination”), being punished (e.g., “I was criticized and punished”), bereavement (e.g., “A family member/close friend died”), change for adaptation (e.g., “My living habits changed”), and others. Responses fall into five levels from 1 point (not at all) to 5 points (very much), and the higher the score, the greater their life pressure. At present, the scale is generally used to measure stress levels of Chinese students. For the present study, the Cronbach’s α of this scale was 0.92. The total scores were categorized into high, medium, and low negative life events groups using a cut-off of 1 standard deviation (SD) above, between, and below the mean.


Loneliness was measured by the University of California Los Angeles Loneliness Scale (UCLA LS) [22], a unidimensional, self-report measure of perceived isolation (e.g., items such as “I often feel that there is nobody who cares about me.”). Items are rated on 5-point scale with higher scores indicating greater loneliness. The UCLA LS shows high internal consistency and adequate convergent validity [23]. Internal consistency in our study was high for the sample (Cronbach’s α = 0.79). The score hierarchies were in accordance with the negative life events scale.


Demographic characteristics

The demographic characteristics of the LBC and NLBC groups are shown in Table 1. From the total sample of 13,952, the NLBC group represented 56.85% (n = 7,918), and the LBC group comprised 43.25% (n = 6,034). In 23.36% of cases (n = 3,257), the father had migrated to an urban area for work; in 3.14% (n = 438) it was the mother, and in 16.72% (n = 3,257) of cases, both parents had migrated. Most students in the sample were taken care of by the mother on a daily basis (68.02%). Most students were from a middle socioeconomic background (54.97%). A small proportion of the children were only children (34.23%). There was no significant difference between the LBC and non-LBC groups in terms of gender. However, we observed large differences between the two groups. Specifically, the LBC group had a mean socioeconomic status that was 1.55% higher than the non-LBC group, the education level of mothers in the LBC group was 11.62% higher, the number of cross-generational families was 5.79% higher, and the rate of being an only child was 14.72% higher. Conversely, a democratic parenting style was 3.39% higher in the non-LBC group, and family history of mental illness was 1.12% higher. Generally, positive factors were higher in the NLBC group than in the LBC group, and vice versa.

Association of demographic characteristics with suicide attempts

The self-reported one-year prevalence rate of SAs was 3.24%. Across the whole sample, girls were more likely to have attempted to commit suicide than were boys, OR 1.22 (95%CI: 1.01–1.47). However, in the LBC group, this difference was not significant. Additionally, significant differences were found between boys in the LBC and non-LBC groups (P < 0.01) in terms of one-year prevalence of SAs. Parenting style was significantly associated with SAs. This association was further replicated in multiple regression analysis when controlling for age, gender, abuse, and neglect. Additionally, family history of mental illness was associated with an increase in the risk of SAs, not only in the LBC group, OR 0.33 (95%CI: 0.22–0.51), but also in the non-LBC group, OR 0.37 (95%CI: 0.23–0.60) (shown in Table 2).

Table 2. Association of demographic characteristics with suicide attempts.

Several variables in the subgroup analyses were not significantly different (P > 0.05) across groups, comprising female gender, age between 10 and 12 years old, not only child status, nuclear families, extended families, single-parent families, parents and caregivers with a college or high school degree, and family history of mental illness. However, there were differences between the LBC and non-LBC groups, including the number of children in 12th grade, male gender, no family history of mental illness, parents and caregivers who completed junior middle school, and only child status.

Association of CTSPC, ASRLEC, and UCLA LS scales with suicide attempts

Importantly, 66.97% of the children in rural China had experienced neglect. Nearly half of the students (46.77%) had experienced abuse categorized as mild, moderate, or severe. In total, 72.66% of children reported having a “negative life” at a level classified as moderate or above, and 69.90% said that they had experienced loneliness, again to a degree categorized as moderate or above. The proportion of LBC experiencing a high level of loneliness was 3.75%, higher than the non-LBC group (χ2 = 11.58, P < 0.01). Overall, the three scales were significantly positively associated with SAs. Considering the difference between the LBC and non-LBC groups, the prevalence of SAs was higher in the LBC group than in the non-LBC group in terms of those children scoring on the middle level of the abuse scale (χ2 = 5.55, P < 0.05) and high in terms of loneliness (χ2 = 11.90, P < 0.05). There were no significant differences between the LBC and non-LBC groups at other levels (shown in Table 3).

Table 3. Association of CTSPC, ASRLEC, and UCLA LS scales with suicide attempt.

Association of LBC characteristics with suicide attempts

The risk of SA in children for whom both parents had migrated was significantly higher than in children for whom either the mother or the father had migrated for work, OR 1.39 (95%CI: 1.02–1.78). Moreover, the risk of SAs was significantly higher in children who communicated with their parent(s) every six months or more compared with those who communicated with a frequency of more than once per week, OR 2.39 (95%CI: 1.59–4.36). For children whose parents had migrated to urban areas for work when they were between 7 and 12 years old, the risk of SAs was reduced compared with those whose parents had left when they were under 2 years old, OR 0.62 (95%CI: 0.43–0.89). However, there were no significant differences across groups in terms of length of time since parents had migrated, frequency of reunions with parents, and length of parents’ stay at home during each reunion (shown in Table 4).

Table 4. Association of LBC characteristic information and suicide attempts.

Logistic regression of potential predictors of suicide attempts

Multivariate logistic regression was performed to examine the influence of certain predictors on SAs (as summarized in Table 5). After adjusting for socioeconomic and demographic characteristics, neglect, physical abuse, negative life events, and loneliness were strongly associated with risk of SAs in both LBC and NLBC groups. Gender was a significant variable in the single factor analysis but did not enter the regression equation in the total sample. Whereas gender was a significant predictor in the NLBC group, OR 1.43 (95%CI: 1.17–2.16), it was not in the LBC group. Conversely, whether either parent had remarried was an important predictor in the LBC group but not in the NLBC group, OR 3.55 (95%CI: 1.49–6.57).

Table 5. Results of logistic regression analyses predicting suicide attempts of NLBC and LBC groups.

Overall, the higher the family’s socioeconomic status, the greater the risk of suicide, OR 0.82 (95%CI: 0.69–0.97), but this was not found in either subgroup on its own. Family history of mental illness was a predictor within the total sample, OR 1.72 (95%CI: 1.1–92.63) but, again, was not found uniquely in either subgroup. All the equations indicated that risk of SAs lowered as age increased.


This quantitative survey-based study presents cross-sectional evidence suggesting that LBC are disadvantaged in terms of their families’ socioeconomic situations and parenting styles. As described in previous literature, parental migration from rural to urban areas often leads to a higher income and enhanced socioeconomic status [24]. However, reduced parental care in the absence of one or both parents negatively influences a child’s development and probably also harms a child’s psychological and physical health due to lessened family control and supervision and weakened parental support and guidance combined with undermined parent—child bonding. This combination of adversities could feasibly manifest in a higher proportion of children being left behind. Given the differences between LBC and NLBC in demographic characteristics, it is not difficult to explain why prevalence of SAs was higher in the LBC group than in the NLBC group.

Prevalence of SAs in rural China was 3.24%, far less than estimates reported in Western societies [25,26] but similar to reports in China [27]. The increased likelihood of girls having attempted suicide relative to boys is consistent with many previous studies [28,29]. Previous research has suggested that girls are more sensitive to interpersonal relationships, including those with peers and family. They were also more inclined than boys to hide negative emotions [30]. The huge difference in suicidal behaviors between urban and rural areas is an important characteristic in China [31]. However, few differences were found between urban and rural areas in the present study (3.5% vs. 3.24%)—a departure from what we found previously [27]. When interpreting these estimates of SAs prevalence, it is important to note that there may be unaccounted-for factors that could cause inaccurate self-reporting. In this study, we asked the students to write their student numbers on the questionnaires so that they could be identified for further research. It is possible that the students may have been hesitant or reluctant to give information on a written survey pertaining to such a sensitive topic as suicidal behavior. Additionally, social desirability bias should be considered. With these limitations in mind, the prevalence of suicide in rural China, particularly in LBC, still deserves serious attention.

Interestingly, in single factor analysis, the higher the mother’s education level, the higher a child’s suicide rate in high school or technical secondary school as well as in college or higher education. Moreover, the prevalence of SAs was highest in children whose caregivers had been educated to a college degree or above. Similar associations were not found in relation to the father’s level of education. This was found across both groups of children, a finding that has not been commonly reported. The critical emphasis on education is a unique characteristic of the Chinese culture [29,32,33]. The putative explanation is that highly educated parents will have high expectations of their children, which, in turn, could exacerbate academic pressure. The novel findings suggest that further studies should be conducted to help us understand the relationship between academic stress and risk behavior, indexed by SAs, in China. It is important to note that, the conclusion requires further validation because these data are based on retrospective self-reports of the occurrence of SAs rather than actual attempt, which introduces potential problems with reporting bias leading to false associations. Due to reporting bias, it is difficult to know whether these reported suicide attempts are really correlated with actual suicides. It is possible that those with better family background report more suicide attempts whereas those with worse family background are more likely to commit suicide, which needs further research and demonstration. In addition, the reliability of the inference may be limited according to the small sample size of subgroup.

Despite differences in the demographics of our LBC and NLBC, negative parenting styles lead to SAs in both groups. This finding highlights the salient role of parenting style, and previous empirical evidence can shed light on this matter [34]. Furthermore, there is an accumulating body of literature suggesting that children with remarried parents are at increased risk of self-harm and SAs. These children must face a more complicated living situation, requiring communication with their original family as well as step families [29,35]. Our findings support the idea that LBC from remarried families were more at risk for attempting suicide. However, NLBC from cross-generational families were more at risk for attempting suicide.

Compared with Western counterparts, Chinese children endure more neglect and physical abuse. In our study, the prevalence of neglect and physical abuse was 66.97% and 46.77%, respectively. However, it was estimated that the prevalence of physical abuse in Australian children ranges from 5% to 18% whereas neglect ranges from 2% to 12% [36]. Our findings show that neglect and physical abuse is also an important issue in rural China. In the current study, neglect and physical abuse were both risk factors for SAs in both the total sample and subgroups. Previous research has found that adolescents who experience neglect and physical abuse may be at risk for SAs [37]. Indeed, not all LBC are necessarily neglected, but parents should communicate with their children more frequently. In the current study, we found that higher communication frequency between parents and children was associated with lower SAs.

The strongest predictor of SAs among the three scales was negative life events. This conclusion was consistent with evidence of a correlation between negative life events in the preceding year and SAs [38]. In this vein, it is likely that children already exposed to negative life events such as failing a test, stress from interpersonal relationships, and pressure to enter a better high school or college should receive timely psychological support and encouragement from their parents or guardians. Researchers found that resilience could be promoted through parental monitoring, disclosure, social support, and family-based interventions [39]. Further research is required to identify protective factors against students’ suicidality within rural China.

The present study indicates that loneliness was the most common and important experience of both groups. Children who experienced high levels of loneliness had higher levels of SAs in the LBC group than in the NLBC group. This is an important finding. In the present study, we found that the risk of SAs in children who had been left behind by both parents was significantly lower than in children for whom only one parent had left. Parents who leave home should be aware of the prominence of their involvement in their children’s lives even if they cannot physically accompany them. They should communicate closely with their children via telecommunication to reduce the loneliness of LBC.

In this comparative study, compared with families with an optimal socioeconomic status, the general group had a lower risk of SAs. This was only found in the NLBC and not in LBC. Nonetheless, there were no statistical differences between the optimal and low socioeconomic status groups. In contrast, according to the existing literature, family economic adversity significantly affects adolescents’ suicidal behaviors [29,40]. Parents in families with a high socioeconomic status may work more often than those from general levels, which may lead to neglect or a spoiled parenting style in the context of rural China. If this conjecture is correct, we must study the relationship between family economic status and neglect of children in rural China. Furthermore, we identified age as a protective factor; older LBCs had a smaller risk for SAs.

Our novel findings suggest that more comparative studies are required to explore the role of parental migration to urban areas for work in negative outcomes in LBC. The unique characteristics of LBC and NLBC in China suggest a pressing need to investigate risk factors and to determine the factors that might be targeted in intervention programs. This study is relevant to future international studies as well as informing culturally based prevention and intervention programs and services targeted toward children in rural China.

This study should be interpreted in light of certain limitations. First, these data are based on retrospective self-reports of the occurrence of SAs rather than actual attempt, which introduces potential problems with reporting bias leading to false associations, so the conclusion requires further validation. Second, although the study achieved a relatively large sample size, there were some demographic differences between the LBC and NLBC, which could indicate selection bias. The current study was a school-based sample, and LBC may drop out of school before finishing compulsory education and before their NLBC counterparts [41]. Further studies should incorporate a group of children no longer attending school to minimize this bias. Third, these data are based on retrospective self-reported data of the occurrence of suicidal behaviors in the past 12 months, which may be underestimated because of recall bias, introducing potential problems with under-reporting in terms of pseudo-anonymity and socioeconomic background. Finally, this study was cross-sectional in nature, meaning that no causal inferences can be made. Thus, more longitudinal research is needed to understand the mechanism of how risk factors lead to SAs on individual and group levels.


Our findings confirmed that SA prevalence among LBC was higher than among NLBC in rural China. Neglect, physical abuse, negative life events, loneliness, and parenting style were risk factors for SAs while more communication with LBC by guardians may reduce the burden of suicide in this group. However, further research is needed to explain how risk factors play a role at both the psychological, social, and family background levels, which would benefit suicide intervention and prevention policies in rural China and worldwide. SAs present an important public health issue in rural China, and more attention should be paid to suicide prevention among students from this area, especially those who have been left behind.

Supporting information

S1 Table. Raw data of a comparative analysis of suicide attempts in left-behind children and non-left-behind children in rural China analysis.

Data are from the nationwide study on mental health outcomes among adolescents in rural China whose authors may be contacted at the Department of Child, Adolescence and Woman Health Care, School of Public Health, Tongji Medical College, Huazhong University of Science and Technology, 430030,Wuhan, Peoples Republic of China.



The authors importantly thank all the participants and their families.

Author Contributions

  1. Conceptualization: HC YY.
  2. Funding acquisition: YY.
  3. Investigation: JH YM XY.
  4. Project administration: YY.
  5. Resources: YY.
  6. Software: QY.
  7. Supervision: YY.
  8. Visualization: YY.
  9. Writing – original draft: HC QY.
  10. Writing – review & editing: LT.


  1. 1. Ougrin D, Tranah T, Stahl D, Moran P, Asarnow JR (2015) Therapeutic interventions for suicide attempts and self-harm in adolescents: systematic review and meta-analysis. J Am Acad Child Adolesc Psychiatry 54: 97–107.e102. pmid:25617250
  2. 2. Sampasa-Kanyinga H, Dupuis LC, Ray R (2015) Prevalence and correlates of suicidal ideation and attempts among children and adolescents. Int J Adolesc Med Health.
  3. 3. Kuo WH, Gallo JJ (2005) Completed suicide after a suicide attempt. Am J Psychiatry 162: 633.
  4. 4. Cluver L, Orkin M, Boyes ME, Sherr L (2015) Child and Adolescent Suicide Attempts, Suicidal Behavior, and Adverse Childhood Experiences in South Africa: A Prospective Study. J Adolesc Health 57: 52–59. pmid:25936843
  5. 5. Brown D (2009) Exposure to physical and sexual violence and adverse health behaviours in African children: results from the Global School-based Student Health Survey. Bulletin of the World Health Organization 87: 447–455. pmid:19565123
  6. 6. Zhao J, Liu X, Wang M (2015) Parent-child cohesion, friend companionship and left-behind children's emotional adaptation in rural China. Child Abuse Negl 48: 190–199. pmid:26190190
  7. 7. Hoertel N, Franco S, Wall MM, Oquendo MA, Wang S, et al. (2015) Childhood maltreatment and risk of suicide attempt: a nationally representative study. J Clin Psychiatry 76: 916–923; quiz 923. pmid:26231006
  8. 8. Givaudan M, Pick S (2013) Children left behind: how to mitigate the effects and facilitate emotional and psychosocial development: supportive community networks can diminish the negative effects of parental migration. Child Abuse Negl 37: 1080–1090. pmid:24268378
  9. 9. Ji K, Finkelhor D (2015) A meta-analysis of child physical abuse prevalence in China. Child Abuse Negl 43: 61–72. pmid:25498804
  10. 10. Seth R (2015) Child Abuse and Neglect in India. Indian J Pediatr 82: 707–714. pmid:25465678
  11. 11. Young R, Lennie S, Minnis H (2011) Children's perceptions of parental emotional neglect and control and psychopathology. J Child Psychol Psychiatry 52: 889–897. pmid:21438874
  12. 12. Linda WP, Marroquin B, Miranda R (2012) Active and passive problem solving as moderators of the relation between negative life event stress and suicidal ideation among suicide attempters and non-attempters. Arch Suicide Res 16: 183–197. pmid:22852781
  13. 13. Zhang WC, Jia CX, Zhang JY, Wang LL, Liu XC (2015) Negative life events and attempted suicide in rural China. PLoS One 10: e0116634. pmid:25611854
  14. 14. Schinka KC, van Dulmen MH, Mata AD, Bossarte R, Swahn M (2013) Psychosocial predictors and outcomes of loneliness trajectories from childhood to early adolescence. J Adolesc 36: 1251–1260. pmid:24007942
  15. 15. Page RM, Yanagishita J, Suwanteerangkul J, Zarco EP, Mei-Lee C, et al. (2006) Hopelessness and Loneliness Among Suicide Attempters in School-Based Samples of Taiwanese, Philippine and Thai Adolescents. School Psychology International 27: 583–598.
  16. 16. Su S, Li X, Lin D, Xu X, Zhu M (2013) Psychological adjustment among left-behind children in rural China: the role of parental migration and parent-child communication. Child Care Health Dev 39: 162–170. pmid:22708901
  17. 17. Mo X, Xu L, Luo H, Wang X, Zhang F, et al. (2015) Do different parenting patterns impact the health and physical growth of 'left-behind' preschool-aged children? A cross-sectional study in rural China. Eur J Public Health.
  18. 18. Wen M, Lin D (2012) Child development in rural China: children left behind by their migrant parents and children of nonmigrant families. Child Dev 83: 120–136. pmid:22181046
  19. 19. Cortes P (2015) The Feminization of International Migration and its Effects on the Children Left Behind: Evidence from the Philippines. World Development 65: 62–78.
  20. 20. Straus MA, Hamby SL, Finkelhor D, Moore DW, Runyan D (1998) Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: development and psychometric data for a national sample of American parents. Child Abuse Negl 22: 249–270. pmid:9589178
  21. 21. Peng L, Zhang J, Li M, Li P, Zhang Y, et al. (2012) Negative life events and mental health of Chinese medical students: the effect of resilience, personality and social support. Psychiatry Res 196: 138–141. pmid:22405636
  22. 22. Russell DW (1996) UCLA Loneliness Scale (Version 3): reliability, validity, and factor structure. J Pers Assess 66: 20–40. pmid:8576833
  23. 23. Hirsch JK, Chang EC, Jeglic EL (2012) Social problem solving and suicidal behavior: ethnic differences in the moderating effects of loneliness and life stress. Arch Suicide Res 16: 303–315. pmid:23137220
  24. 24. Zhao X, Chen J, Chen MC, Lv XL, Jiang YH, et al. (2014) Left-behind children in rural China experience higher levels of anxiety and poorer living conditions. Acta Paediatr 103: 665–670. pmid:24527673
  25. 25. Peltzer K, Pengpid S (2015) Early Substance Use Initiation and Suicide Ideation and Attempts among School-Aged Adolescents in Four Pacific Island Countries in Oceania. Int J Environ Res Public Health 12: 12291–12303. pmid:26437423
  26. 26. Hauser M, Galling B, Correll CU (2013) Suicidal ideation and suicide attempts in children and adolescents with bipolar disorder: a systematic review of prevalence and incidence rates, correlates, and targeted interventions. Bipolar Disord 15: 507–523. pmid:23829436
  27. 27. Tang J, Yu Y, Wu Y, Du Y, Ma Y, et al. (2011) Association between non-suicidal self-injuries and suicide attempts in Chinese adolescents and college students: a cross-section study. PLoS One 6: e17977. pmid:21494656
  28. 28. Law BM, Shek DT (2013) Self-harm and suicide attempts among young Chinese adolescents in Hong Kong: prevalence, correlates, and changes. J Pediatr Adolesc Gynecol 26: S26–32. pmid:23683824
  29. 29. Law BMF, Shek DTL (2015) A 6-year Longitudinal Study of Self-harm and Suicidal Behaviors among Chinese Adolescents in Hong Kong. Journal of Pediatric and Adolescent Gynecology.
  30. 30. Ellis JB, Lamis DA (2007) Adaptive characteristics and suicidal behavior: a gender comparison of young adults. Death Stud 31: 845–854. pmid:17886414
  31. 31. Hashimoto K, Yang L-S, Zhang Z-H, Sun L, Sun Y-H, et al. (2015) Prevalence of Suicide Attempts among College Students in China: A Meta-Analysis. Plos One 10: e0116303. pmid:25664661
  32. 32. Yadegarfard M, Ho R, Bahramabadian F (2013) Influences on loneliness, depression, sexual-risk behaviour and suicidal ideation among Thai transgender youth. Cult Health Sex 15: 726–737. pmid:23659441
  33. 33. Liu X, Tein JY, Zhao Z, Sandler IN (2005) Suicidality and correlates among rural adolescents of China. J Adolesc Health 37: 443–451. pmid:16310121
  34. 34. Greening L, Stoppelbein L, Luebbe A (2010) The moderating effects of parenting styles on African-American and Caucasian children's suicidal behaviors. J Youth Adolesc 39: 357–369. pmid:19806443
  35. 35. Lee K, Namkoong K, Choi W-J, Park JY (2014) The relationship between parental marital status and suicidal ideation and attempts by gender in adolescents: Results from a nationally representative Korean sample. Comprehensive Psychiatry 55: 1093–1099. pmid:24746529
  36. 36. Sun J, Buys NJ (2013) Child abuse, neglect and maltreatment health service in Australia: A literature review.
  37. 37. Norman RE, Byambaa M, De R, Butchart A, Scott J, et al. (2012) The long-term health consequences of child physical abuse, emotional abuse, and neglect: a systematic review and meta-analysis. PLoS Med 9: e1001349. pmid:23209385
  38. 38. Donath C, Graessel E, Baier D, Bleich S, Hillemacher T (2014) Is parenting style a predictor of suicide attempts in a representative sample of adolescents? BMC Pediatr 14: 113. pmid:24766881
  39. 39. Eloff I, Finestone M, Makin JD, Boeving-Allen A, Visser M, et al. (2014) A randomized clinical trial of an intervention to promote resilience in young children of HIV-positive mothers in South Africa. AIDS 28 Suppl 3: S347–357.
  40. 40. Aschan L, Goodwin L, Cross S, Moran P, Hotopf M, et al. (2013) Suicidal behaviours in South East London: prevalence, risk factors and the role of socio-economic status. J Affect Disord 150: 441–449. pmid:23726782
  41. 41. Li Q, Zang W, An L (2013) Peer effects and school dropout in rural China. China Economic Review 27: 238–248.