Introduction
China’s reform and opening-up and the subsequent social and economic development have led to the prosperity of many cities. A large amount of the population in the undeveloped areas in central and western China began to move to the developed area for employment opportunities. According to the Report on China’s Migrant Population Development, there were more than 200 million floating populations in China in 2020, and over 70% of them gathered in the developed region in eastern China. The data show that the average monthly income of the floating population is 4598 yuan. After paying for food, clothing, housing, and transportation, their disposable income is too low to afford their children’s education and housing in the developed region. Therefore, a large number of children are left behind by their parents in their hometown, who are known as left-behind children. The left-behind children are defined as children who have lived in undeveloped areas for more than 6 months and have one parent or both work in developed areas. In 2015, there were 68.77 million left-behind children in China, including 40.51 million in rural areas [
1]. These children are usually taken care of by their relatives, especially grandparents, who pay more attention to their physiological needs rather than their psychological needs.
Although studies have shown that working in the developed area can increase family income and raise family status in the local area, it is rarely beneficial for left-behind children [
2]. Researches show that left-behind children have more mental health problems than non-left-behind children [
3‐
6], including a higher risk of depression [
7], and anxiety [
8], a stronger sense of loneliness [
9], and higher suicide risk [
10], highlighting severe psychological problems of these children and tough challenges posed by the problems.
Depression is the most common mental health problemin childhood [
11], and depression in children and adolescents is usually the first episode of depression [
12]. Early prevention and intervention can help reduce depression symptoms and the risk of recurrence [
11,
13]. A previous systematic review and meta-analysis showed that left-behind children have higher depression risk and depression scores than non-left-behind children (RR 1.52 [95% CI 1.27–1.82]; SMD 0·16 [0·10–0·21]) [
3]. However, at present, most studies only compare the differences in the score of depression scale between left-behind and non-left-behind children [
14], while few studies have made a detailed comparison of depressive symptoms between left-behind and non-left-behind children. Therefore, it remains unknown whether the symptoms of depression differ between them.
Depression of children is associated with many factors, such as environmental stress [
15], life trauma [
16], bullying [
17], etc. Studies show that children who are exposed to negative life events are at a higher risk of depression [
18]. It is also reported that left-behind children are more likely to be exposed to negative life events. For example, in rural areas of Mexico, the children whose father works in other places have 39% higher incidence of disease than the non-left-behind children, and their incidence of diarrhea is increased by 51% [
19]. In addition, the risk of physical and mental disorders in the left-behind children is higher than that in the non-left-behind children, such as car accidents, falls [
3], sexual abuse, neglect [
20] and bullying [
21], and less communication with parents [
6]. It is worth noting whether the negative life events experienced by left-behind children aggravate their depressive symptoms. Studies have shown that reducing these negative life events can alleviate depressive symptoms. For example, more frequent and longer parent-child communication can significantly reduce the incidence of depression in left-behind children [
18]. Therefore, it is significant to investigate the effect of negative life events on the depression of the left-behind children.
We found in our previous studies that the overall prevalence of depression in the left-behind children was 24.8%. Meanwhile, we also found that high income, frequent parent-child communication, telephone communication or talking about learning experiences, school life and emotional experiences are protective factors for depression [
22]. In addition, prevalence of depressive symptoms is higher in the left-behind children than in non-left-behind children, and negative life events are risk factors for depression in left-behind children [
18]. Despite the work we have done above, it is still unclear which negative life event plays a major role in certain type of depressive symptom.
Few studies have investigated the relationship between negative life events and depression symptoms of left-behind children, and to our knowledge, no studies have been conducted to explore their relationship using network analysis. Therefore, the understanding remains limited regarding their relationship, as well as the major types of the negative life events experienced by the left-behind children and their core depression symptoms. This study aims to explore the relationship between negative life events and characteristics of depression symptoms of left-behind children using network analysis. The network usually consists of symptoms (nodes) of mental disorders and connections (edges) between them [
23]. According to the network theory, diseases are usually caused by the interaction between symptoms, and the symptoms themselves are not mental diseases, but components of mental diseases [
24]. Alleviating the symptoms of disease plays an important role in disease treatment. Targeting the core symptom derived from network analysis helps us treat the underlying disease [
25].
We investigated the network structure of depressive symptoms of a large sample (N = 2517) of left-behind children and explored the relationship between depression and negative life events using network analysis. We aimed to: a) construct depression networks to identify the main symptoms of depressive disorders of left-behind children; b) compare the differences in depression networks between left-behind children and non-left-behind children; and c) use bridge centrality to identify the disease pathway linking negative life events and depressive symptoms of left-behind children.
In this study, two networks were constructed: the network of depressive symptoms and the network of both depressive symptoms and negative life events. These networks were used to identify the most central symptoms of depression disorders of left-behind children, to identify the negative life events with a stronger correlation with depressive symptoms of left-behind children using bridge symptoms, and to compare the differences in the two networks between left-behind children and non-left-behind children.
Discussion
We used network analysis to study the characteristics of depression network and depression-negative life events network with a large sample (
N = 2517) of left-behind children. We found that differences existed in the strength of symptoms in the CDI network of left-behind children, and self-hatred has the highest centrality in the network, consistent with the previous research which shows that self-hatred also has the highest centrality in the CDI of the non-left-behind children [
42]. This means that left-behind and non-left-behind children have similar central symptoms in the CDI depression network. NCT analysis in our study also shows that there is no significant difference in the CDI depression network between left-behind and non-left-behind children. These findings may be explained by the fact that left-behind and non-left-behind children share the same developmental stage, such as puberty. On the one hand, the left-behind children begin to have self-awareness [
43], and the derived self-related internal information [
44] plays an important role in depression [
45]. On the other hand, with the development of self-identity in adolescence, the children tend to pay more attention to self-achievements, family atmosphere, etc. [
46]. Indeed, previous studies have shown that lower self-worth is associated with depression [
47]. The negative self-information has become a potential risk factor of adolescent depression [
48], while positive self-identity information can prevent negative effects [
49]. Last but not least, depression may also be associated with physical development during puberty, such as the development of secondary sexual characteristics and height [
50], and with excessive attention paid to appearance [
51].
In addition, we also found some other high-intensity symptoms, such as crying, fatigue and sadness. Among them, crying and fatigue are not the most central symptoms according to the previous report [
42]. Our finding about fatigue and sadness can be explained by the fact that left-behind children experience more bullying or abuse [
52], and they can only reduce the risk of depression through self-sympathy [
21]. Fatigue symptoms stem from more manual labor they need to shoulder when living with grandparents. Compared with non-left-behind children, left-behind ones reported increased working hours [
53], and the fatigue increased the risk of depression [
54,
55]. Therefore, we should improve the social support for left-behind children, including protecting them against abuse and bullying and reducing their extra working hours, to reduce their risk of depression.
It is interesting to find that these central symptoms in the depression network of the left-behind children differed from those in CDI-ASLEC network which included the negative life events. In CDI-ASLEC network, academic stress and public humiliation in ASLEC and school performance decrement in CDI have the highest centrality reflecting great academic stress and impaired self-esteem of the left-behind children. It should be noted that these networks are undirected rather than directed (causal networks). The most central symptoms were not obtained based on all symptoms, but on the symptoms we input.
We also identified bridge symptoms in CDI-ASLEC network. Bridge symptoms are considered to be an illness pathway for one disorder to spread to another. Therefore, when one disorder appears, intervention of potential bridge symptoms can effectively prevent the spread of disorders and the development of complications [
39]. In this network, academic stress, discrimination and school performance decrement have the strongest bridge strength centrality in the left-behind children, consistent with the previous studies. Parents’ inquiry about study accounts for a high proportion of their communication with their children [
18], which increases the children’s academic stress, and in turn increases the risk of depression [
56]. In addition, the self-esteem of left-behind children will be impaired by discrimination by others [
57], and low self-esteem is more likely to lead to depression [
58]. Taking corresponding measures to intervene bridge symptoms, especially those with high intensity centrality, can effectively reduce the risk of depression. For example, some studies have reported that regulating self-esteem bridge symptoms can reduce the impact of negative life stress on depression [
57]. However, future studies are warranted to confirm if academic pressure and difficulties can serve as targets for intervention.
Finally, we used the network comparison test to compare the differences between left-behind children and non-left-behind children in CDI network and CDI-ASLEC network. It is found that, firstly, in CDI network, there is no difference in network structure and global strength between left behind and non-left behind children; secondly, in CDI-ASLEC network, there is no difference in network structure between the left-behind and non-left-behind children. However, there is difference in global strength, indicating that the characteristics of left-behind children network have stronger connection than those of the non-left-behind children. These results show that compared with non-left-behind children, negative life events have a greater impact on left-behind children. One reason is that left-behind children are exposed to more negative life events; another reason is that left-behind children show more depressive symptoms when they encounter negative life events than non-left-behind children [
18]. Therefore, stress caused by the negative life events is a common risk factor for depression in the left-behind children.
In short, this study uses network analysis to obtain more subtle results. Our previous research found that negative life events are a risk factor for depression, and proper and adequate communication between the left-behind children and their parents can effectively reduce the impact of negative life events on depression [
18]. This study further explored the relationship between negative life events and depression from the perspective of symptoms. The results regarding bridge symptoms indicate that academic stress and school performance decrement are the main pathways linking negative life events and depression. This result further supports our previous findings that the parent-child communications about learning experiences and school life are protective factors for depression [
22]. When communicating with the left-behind children, the parents should avoid increasing their academic stress, and meanwhile, they should show more care about their study and life. By doing so, they can reduce the depression of left-behind children caused by negative life events more effectively.
Limitations
There are some limitations in this study. First, we used sample data from the results of a cross-sectional survey, thus we cannot infer the causality dynamically. Second, the data is from previous research, with a poor timeliness. Therefore, it is necessary to repeat the investigation within a certain period of time to explore the evolvement of depressive symptoms of the left-behind children. In addition, due to the limitations of the study method itself, we did not input all the symptoms into the network model while other surveys, using different survey tools, found network structures differ from those of CDI [
59,
60].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.