According to the 2009 census in China, there were 211 million internal migrants and among them, more than 78% were rural-to-urban migrants [1]. Since the early 1990s, one of vital changes of migration pattern is from individual to family migration [1, 2]. For example, from 1993-1997, the size of the floating population remained stable in Shanghai, but the number of migrant children of school age increased from 280,000-340, 000 [3]. At present, approximately 20.8% of Chinese internal migrants are school-aged children (not above 14 years old), equivalent to approximately 43.89 million individuals who have accompanied their parents in relocating to cities [1].

Migration and migration-related processes have been widely assumed to affect the health of children and adolescents [4], especially rural-to-urban migration [5, 6]. For example, vaccination and regular health examination rates among migrant children and adolescents are reported to be significantly lower than the average for their urban peers [7]. In addition to lack of health insurance, access to care may be limited by parental lack of knowledge of health care needs and resources [7, 8]. Studies that focus on the mental health status of Chinese migrant children and adolescents generally show disadvantages, such as lower self-esteem [9]; higher loneliness, social anxiety, and perceived discrimination [10]; and more behavioral problems [11].

Prior reports indicate that low socioeconomic status (SES), social exclusion, and fragmentation of social support networks may lead to poor self-esteem, depression, and anxiety in immigrant populations [12–14]. The relationship between social connections and individual well-being has long been recognized in Western literature. It dates back to the work of Durkheim [15], who documented a link between social isolation and reduced psychological well-being. Since then, a wide range of studies have identified beneficial effects from social networks (including family, friends, neighbors), and social cohesion on individual health [16–18]. Zhang and colleagues [18] identified three layers of social connections-the inner layer of family cohesion, the intermediate layer of support by friends and relatives, and the outer layer of social cohesion. Several surveys report that these layers are differentially related to better health, and may have significant effects on migrants' health [16, 18]. For instance, family cohesion has independent and direct effects on both self-rated physical and mental health among Asian Americans. However, the impact of other social connection measures is partially mediated by SES and immigration-related factors [18].

A large body of evidence suggests that important relationships for adolescents usually fall into the following domains: family, peer group, school, and the community and society at large [19, 20]. The quality of family relationships is central to development during childhood and adolescence. Harker [21] found that migrant children and adolescents who felt closer to their parents tended to score higher on various measures of psychological adjustment, and suffer few adverse effects from stressful events. Data indicate that relationships with peers are also one of the most important factors in the social adaptation of migrant adolescents [22]. In the school domain, a teacher-student relationship is very influential in the development of school-aged children and adolescents. In addition, schools can strengthen social and cultural capital, especially among immigrant pupils [23]. In the social cohesion domain, a social network and perceived social support are also related to adolescents' inner and outer problems, loneliness, school adaptation [24], and self-esteem [9, 25].

Based on prior research [18–20], we differentiated adolescents' social connections into four types-family cohesion, peer association, teacher-student relations, and social cohesion. However, there are few studies on the relation of Chinese migrant adolescents' multiple social connections and health. Most often, reports focus largely on migrants' limited resources, e.g., poor living conditions, limited access to urban health care, and vulnerability to infectious diseases, including HIV and STDs [26, 27]. To date, scant data exist on the health effects of migration on Chinese adolescents. In addition, little empirical work has simultaneously examined multiple social connections in relation to the health of migrant adolescents.

The aims of this study are to compare the self-rated physical and mental health of migrant adolescents and native peers; examine the effects of different types of social connections on adolescents' self-rated health; explore whether the effects of layered social connections vary according to an individual's migrant status; and examine whether SES and demographic factors (e.g., gender, family income, parental education) affect the health of migrants differently than that of local adolescents.

Our study showed that original residents gained a significant advantage from SES, which is consistent with the prior research [5–7]. The three indices of migrant SES were worse than that of natives (P < 0.001). This may be largely due to the substantial proportion of rural-to-urban migrants in our study, whose SES and personal resource were generally poor. Perhaps as a consequence, migrant adolescents showed worse social connections than their local peers, with significant disparities in social cohesion, family cohesion, and peer associations. These outcomes are consistent with those from prior studies in China [6, 24, 25]. Although we did not find a significant difference in student-teacher relations between migrant and native adolescents, local boys reported the lowest scores in teacher relations. A plausible explanation is that migrants and girls are generally considered more likely to obey rules, and respect teachers compared with local students and boys [36].

Our data support prior research showing that migrants have significantly lower self-esteem than local adolescents [9]. The present study also shows that all indicators of SES and social connection indices were positively correlated with self-esteem scores in migrants. The effects of SES among local students were not as significant as migrants. Namely, poor social connections and low family SES may affect the self-confidence and self-esteem of migrant adolescents particularly. Differences in the values and patterns of behavior between the children of migrant workers in China and their urban peers (e.g., heavy regional accent, out-of-date clothing, and communication difficulties) might lead to social discrimination that contributes to a lack of self-confidence and social isolation among migrant children and adolescents [25, 39].

Age was significantly associated with depression among migrants. This outcome is consistent with previous research that identified age as a risk factor for psychological well-being in migrant children and adolescents [6]. Although migrant adolescents showed more depressive symptoms than local youth, the difference diminished when sociodemographic factors were considered. However, the difference in self-esteem between the two samples still persisted. Thus, our data partially support prior research showing that the internal migrant adolescents have poorer psychological health than their local peers in China.

Interestingly, migrant adolescents' self-rated physical health was generally better than that of local peers. This is consistent with the healthy migrant effect, which has been described in many Western countries [40]. Evidence from Chinese adult migrants also demonstrates the healthy migrant phenomenon [17, 41]. One explanation is the health selection hypothesis, which emphasizes that healthier individuals are more likely to migrate [40]. Another hypothesis is that the migrants are more likely to return to their places of origin when they have a serious illness [42]. It is noted that girls are more likely than boys to report fair/poor physical health. This result might reflect differences between migrants and native residents, and boys and girls in their appraisals of health and attitudes about ill health.

Self-rated poor health by migrant adolescents was related to the father's education, i.e., middle or high level. In this study, a middle education level had a protective effect. Miao [43] suggested that high parental education level might be a protective factor for parents themselves, but a risk factor for the health of migrant adolescents. Most likely, parents with a high education level expect academic success from their children; doing poorly in school could cause families to lose face. In China, academic success is a high priority for parents and children. Parents, in turn, pursue all options to ensure that their children work hard and strive for academic excellence [44]. In rural areas, fathers generally have more education than their wives, and are more likely to supervise their children's academic achievement. The outcome might be more academic pressure accompanied by a high level of distress in their offspring. The parental education effect may differ between migrant and local families. In the modern Chinese family, well-educated mothers who become stricter and stricter with their children--"Chinese tiger moms"-- are very popular in urban areas [45]. Our data suggest that the self-rated physical health of adolescents is a complex and multifaceted phenomenon. The results of our study merit further exploration.

Collectively, this report showed that layered social connections are linked to higher self-esteem and lower levels of depression among migrants. This suggests the importance of social connections to mental health in adolescents. We found a stronger sense of peer association and social cohesion to be particularly beneficial for migrants in all three health outcomes. This suggests that the two factors may be particularly protective for immigrants. Migration is characterized by a loss of friendships and challenges to the establishment of new ties. Those who form and maintain good peer relationships and feel more connected with others are likely to have better psychological well-being [25]. Strong social cohesion may mitigate the effects of a lack of other socioeconomic resources for certain populations [17]. Family cohesion was associated with both depressive symptoms and self-esteem among all adolescents. This finding confirms the well-studied importance of family relations for adolescent mental health in immigrant and non-immigrant groups [46–48].

Our results should be interpreted with caution due to methodological limitations of the study. Its cross-sectional design preempts the ability to establish cause-effect relationships. The sample of migrants all came from the suburban public schools in Shanghai, which have a mainly middle-low SES population, specifically among rural-to-urban migrants. That both migrant and local residents were recruited from a single major metropolitan area limits the generalizability of the results. Self-reported demographic and health data might be biased. This is especially true on measures of self-rated physical health [49, 50], which might be subject to reverse causation.

Limitations notwithstanding, this study contributes to the literature by examining the relationships between layered social connections and health issues among Chinese migrant and local adolescents that have been neglected in prior studies. Data show that migrants have a significant disadvantage in psychological well-being and social connections; that family cohesion is an important factor for the psychological well-being of all adolescents; and that social cohesion and peer associations play a strong role, especially as they relate to migrant health. These findings indicate the importance of strengthening the interpersonal systems of adolescents. In particular, social cohesion and positive connections with parents and peers are significant protective factors that can enhance the mental health of migrant adolescents in China. For instance, parents can pay more attention to their children' psychological well-being, and enhance family cohesion. Educators may encourage young migrants to actively engage local peers and organize support groups for those who feel incompetent or inferior to local peers. Such groups should supply information and teach skills that improve interactions with urban residents, and help migrant youth make a more balanced appraisal of their own abilities and qualities [25]. Local organizations can help by organizing social workers and volunteers from nearby universities and colleges, thereby creating a bigger and broader safety net. It may also be important for the government and relevant organizations to study the issue and develop strategies for addressing the needs of migrant adolescents at a local level.