Li et al.

BMC Pediatrics (2020) 20:271

[Link]

RESEARCH ARTICLE Open Access

How do socioeconomic status relate to

social relationships among adolescents: a
school-based study in East China
Jing Li1, Jing Wang2, Jia-yu Li1, Sheng Qian1, Rui-xia Jia3, Ying-quan Wang3, Jing-hong Liang1,3 and Yong Xu1,3*

Abstract
Background: A great number of studies have concentrated on the influence of socioeconomic status with health
outcomes, but little on how socioeconomic status affects social relationship in adolescents’ families, peers and
schools. This study aimed to clarify more detailed information on the connection between social relationships and
different dimensions of socioeconomic status.
Methods: A school-based cross-sectional study was performed by 13–18 adolescents enrolled in East China from
September, 2018 to May, 2019, which recruited 6902 students from junior and senior high schools and used the
stratified random sampling method. Parent-child relationship (cohesion, expressiveness, conflict), peer relationship
(interpersonal relationship, communication and interaction, social emotion) and student-teacher relationship
(intimacy, support, satisfaction, conflict) were investigated. Besides, objective socioeconomic status (parental
education and occupation, assessed by the adolescent) and subjective socioeconomic status (self-evaluation of
family social class) were measured. More detailed information was used to clarify the link between social
relationships and different dimensions of socioeconomic status.
Results: All five indicators of socioeconomic status were slightly positively correlated with the quality of social
relationships (r ranged from 0.036 to 0.189, all p < 0.001), except that maternal education was not correlated with
the conflict dimension of parent-child relationship. Standardized regression coefficients indicated that paternal
education (β = 0.08) and occupation (β = 0.07) were the predictors of parent-child relationship. And peer
relationship model revealed that the corresponding effect size was slightly stronger for subjective socioeconomic
status (β = 0.10), whereas the maternal education had a slightly stronger correlation with student-teacher
relationship (β = 0.07) relative to other indicators.
(Continued on next page)

* Correspondence: xuyong_childhealth@[Link]
1
Department of Child and Adolescent Health, School of Public Health,
Medical College of Soochow University, No.199 Ren Ai Road, Suzhou, Jiangsu
215123, People’s Republic of China
3
Department of Social medicine, School of Public Health, Soochow
University, Suzhou, Jiangsu 215123, People’s Republic of China
Full list of author information is available at the end of the article

© The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License,
which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if
changes were made. The images or other third party material in this article are included in the article's Creative Commons
licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons
licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain
permission directly from the copyright holder. To view a copy of this licence, visit [Link]
The Creative Commons Public Domain Dedication waiver ([Link] applies to the
data made available in this article, unless otherwise stated in a credit line to the data.
Li et al. BMC Pediatrics (2020) 20:271 Page 2 of 10

(Continued from previous page)

Conclusions: Adolescents with lower socioeconomic status had poorer social relationships compared to those with
higher socioeconomic status. These findings have important public health implications for health policy makers to
make sound decisions on resources allocation and services planning in improving adolescents’ social relationships
and promoting health outcomes.
Keywords: Socioeconomic status, Parent-child relationship, Peer relationship, Student-teacher relationship,
Adolescents

Background behavioral communication [19]. Positive student-teacher

Socioeconomic status (SES), as one of the most significant relationship is critical to the positive development of ado-
social determinants of health, has attracted increasing at- lescents. It provides security, safety, and protection that
tention around the globe. Many studies have suggested necessary for students’ full participation in social activities
that education, income and occupation of objective SES [20], and supports them in adjusting to school life, im-
indicators have profound impacts on everyone’s health [1, proving their social skills and promoting academic
2]. Subjective SES is also called as subjective status, per- achievement [21]. Drawing from these perspectives, social
ceived social position [3], and subjective social status [4]. relationship, including parent-child relationship, peer rela-
It is defined as the faith of an individual about his or her tionship and teacher-student relationship, is an important
position in the socioeconomic structure [5]. Some re- factor influencing many health outcomes of adolescents,
searchers demonstrate that subjective SES is more accur- such as health behaviors, mental health, physical health
ate in capturing the more subtle aspects of social status, and death risk [22].
and providing more information than objective SES indi- Whereas a lot of researches have focused on adoles-
cators, it also has a greater impact on health [6]. cents’ health outcomes in terms of socioeconomic status
The effects of SES on health change over the life course or social relationships, few of them know about social
[7]. The importance of socioeconomic factors for infants’ relationships from the perspective of socioeconomic sta-
[8] and adults’ [9] health have been widely demonstrated tus. The current study investigated the family, peer and
to be inverse, and graded correlated. However, there is lit- student-teacher relationships of adolescents in East
tle evidence that adolescents’ health impacted by SES may China to clarify more detailed information on the con-
be consistent [10]. For example, some studies showed that nection between these relationships with different di-
there was an inverse gradient among SES, global health in- mensions of socioeconomic status. We sought to figure
dicators, acute conditions, and health behaviors [11, 12]. out whether subjective SES and objective SES differed in
Meanwhile, some results indicated that SES had no gradi- the strength of adolescents’ social relationships with dif-
ent among non-fatal injury, acute illness, mental health, ferent dimensions.
and self-rated health [13, 14].
Additionally, it is obvious that the ecological environ-
ment and relationships of adolescents’ development have Methods
changed significantly as individuals transfer from child- Participants and procedure
hood to adolescence, however, family, peers and school From September, 2018 to May, 2019, a school-based
are still the most important and direct social context at cross-sectional study was performed by adolescents en-
the microsystem level [15]. A good parent-child relation- rolled in East China. There were 6902 students came
ship is a vital protective factor for adolescents [16]. Peer from middle or high school, including 3355 males and
relationship refers to the relationship which was formed 3547 females. A stratified cluster random sampling
by common activities and mutual contacts between ado- method was performed through four stages of selection
lescents of the same or similar age. Healthy peer relation- (Fig. 1). In the first stage, three administrative regions of
ship is critical to the positive development of adolescents’ Hangzhou, Suzhou and Hefei were selected from Zhe-
cognitive, emotional, social skills, and scholastic adapta- jiang, Jiangsu and Anhui provinces, respectively. Next,
tion, which support them in normative transitions of de- five junior high schools and five senior high schools
velopment [17] and buffer them against the impact of were selected from each of the three regions. Thirdly,
burdensome circumstances in other areas of life [18]. each grade was considered as a single sampling stratum,
Student-teacher relationship is a basic interpersonal con- and two classes were randomly selected from each grade
nection developed by prolonged interaction between stu- of the 30 schools. Finally, all the students in the selected
dents and teachers, which reflects their psychological state classes were taken as study subjects. The inclusion cri-
of seeking satisfaction through emotional, cognitive and teria for the study participants were as follows: (a) aged
13 to 18, (b) all of them lived in Suzhou or Hefei or
Li et al. BMC Pediatrics (2020) 20:271 Page 3 of 10

Fig. 1 Flowchart of sampling and subject recruitment

Hangzhou with their parents for more than 6 months Measurement

prior to the start of the study, and (c) voluntary partici- Sociodemographic variables
pation. Participants with the following characteristics The social and demographic characteristics of the partic-
were excluded: (a) psychological or mental health issues, ipants were recorded in the questionnaires, including the
and (b) absence during the survey. The investigation was age, grade and household registration of the respon-
conducted during the school year, avoiding the school dents. Information from student interviews was used to
examination period. In cooperating schools, teachers assess the family structure based on the person that ado-
were required to inform students about the study and to lescents were living with. The options available were
ensure privacy. Each of these steps were explained in de- father and mother, father and stepmother/mother and
tail by the investigator. All participants filled out the stepfather, father only/mother only, and some other legal
questionnaires anonymously after knowing the purpose guardian. For the analyses, we created a dichotomous
and methods of the survey. In addition, the study was in- variable: intact family (natural parents)/non-intact family
vestigated in the absence of teachers. (stepparent or single parent or other legal guardian).
Previous study has shown that 5.2% of middle school The variable of family size was also obtained through
students in China have a variety of significant mental student interviews, which can be divided into two cat-
health problems [23]. So we estimated the minimum egories (non-only child/only child).
sample size needed for this study based on this preva-
lence. The calculation formula was as follows: Socioeconomic status
( )2 The SES of families was used to reflect the SES of ado-
57:3Z α=2 lescents, which included both subjective SES and object-
n¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
arcsin½εP= P ð1−P Þ ive SES (education and occupation of parents) [24]. All
of indicators were assessed by students, which saved the
In the present study, p was the prevalence of mental cost and time of data collection [25].
health problems (p = 0.052), ɛ was the error tolerance As for education, adolescents were required to indicate
(ɛ = 0.1), alpha (α) was the significance level (α = 0.05, the highest level of education completed by each parent,
Zα/2 = 1.96 accordingly), the estimated minimum sample and we provides the following options: uneducated, pre-
size was 6920. We added an additional 10% to the mini- school, primary school, junior high school, senior high
mum sample size, taking into account the non-response school/ vocational school, college/university, graduate
rate or absence, and targeted at 7612 subjects. A total of school. Then it was recoded as low (≤ ISCED 0, i.e. un-
7326 students participated in the survey. In the end, only educated, preschool), medium (ISCED 1–2, i.e. primary
6902 valid surveys were retained after excluding ques- school, junior high school), and high (ISCED 3–8, i.e. se-
tionnaires with missing items. nior high school/ vocational school, college/university,
Li et al. BMC Pediatrics (2020) 20:271 Page 4 of 10

graduate school) according to the International Standard points with higher scores reflecting a better-quality rela-
Classification of Education (ISCED) [26]. tionship between the adolescents and peers. The total
Students were also asked to answer where their par- measurement yielded an internal consistency (Cron-
ents worked and what job they did. When adolescents bach’s α) of 0.90 in the present study.
were allowed to answer these open-ended questions, the Student-teacher relationship: The students were asked
proportion of non-response or unclassifiable answers as- to describe the quality of their relationship with teachers
sociated with this measurement showed a significant de- using the Chinese version of the Student-Teacher Rela-
crease [27]. Afterwards, based on the International tionship Scale, which was originally developed by Pianta
Standard Classification of Occupations (ISCO-08) [28], [34] and revised by Zhiyong Qu [35]. The Chinese ver-
each parent’s occupation status was recalculated. The oc- sion of this scale has been widely used in middle school
cupational status was divided into three categories based students in China [36]. Students rated their perceptions
on whether the parent had a job, which type of work and of their relationship with the teachers on 23 items, which
what skill level as: low prestige (unemployed), medium assessed the level of support, satisfaction, intimacy, and
prestige (temporary worker, clerks, service and sales, conflict in their relationship. Each item used a 5-point
skilled workers, craft and related, plant/machine opera- Likert scale (definitely does not apply to definitely applies)
tors, elementary occupations, and armed forces), and high and total score was created from the sum of the 23 items
prestige (technicians, professionals and managers). with higher scores reflecting a better-quality relationship
Subjective social status required adolescents to evalu- between the adolescents and teachers. The subscale of
ate the socioeconomic status of their families [29]. It was conflict was scored reversed, so a positive relationship
consisted of a 5-point ladder scale: 1 (the most affluent) characterized by trust, warmth, and low conflict was rep-
- 5 (the most disadvantaged). For the convenience of resented by high scores. The study was reliable and ad-
analysis, we divided the answers into the following three equate with Cronbach’s alpha values of 0.83 for total scale.
categories: lower class (1–2, below average), middle class
(3, average), and upper class (4–5, above average). Statistical analysis
The statistical analysis was performed by using the soft-
Social relationships ware IBM SPSS version 23. A p-value of < 0.05 was con-
Parent-child relationship: The Family Environment Scale sidered statistically significant (two-tailed). Descriptive
(FES) was a self-report tool used to evaluate the social statistics of the presentation of demographic data and
environment of the family [30]. It has been shown to socioeconomic status were performed and presented as
have good reliability and validity in samples of Chinese appropriate frequencies and proportions. The links be-
adolescents [31]. It contained 10 sub-scales and involved tween SES indicators and scores on different dimensions
three broad domains: social relationship within the fam- of social relationships were examined by using Spearman
ily, personal growth or goal orientation, and system correlations analyze. The correlation of r = 0.1 was inter-
maintenance. In this study, we specifically selected sub- preted as small, r = 0.3 was medium, r = 0.5 was large
scales of cohesion, expressiveness, and conflict to evalu- [37]. Furthermore, the correlation of SES with social re-
ate the quality of parent-child relationship. Moreover, lationships were investigated by using hierarchical mul-
this domain was consisted of 27 true or false items, and tiple regression. Adjusted R2 statistics were used to
each sub-scale contain 9 items. The conflict was scored evaluate the overall fit of the model [38]. Standardized
reversed so that three sub-scales were scored in the regression coefficients were computed to evaluate the
same direction [30]. Total score of this domain was 0– relative contributions of these variables to variations in
27 points, and higher score indicated more harmonious social relationships.
of parent-child relationship. The Cronbach’s alpha value
for this sample was 0.71. Results
Peer relationship: The students were asked to describe In this cross-sectional study, there were more junior
the quality of their relationship with peers using the Peer high school students (61.0%) than senior high school
Relationship Scale developed by Wei Yunhua, a Chinese students (39.0%). The mean ages (SD) of junior and se-
scholar [32]. The reliability and validity of the Peer Rela- nior high school students were 13.9 (0.8) and 16.9 (0.8)
tionship Scale have been confirmed in previous studies years, respectively. There were 41.7% of adolescents
on adolescents [33]. This questionnaire was consisted of from urban areas and 58.3% from rural areas. Most of
20 questions, involving three domains: interpersonal re- the students in the sample came from intact families
lationship, communication, interaction, and social emo- (83.6, and 85.9% for junior and senior high school stu-
tion. Each item was assessed on a five-point Likert scale dents, respectively). With regarded to subject SES, a lar-
from definitely does not apply (1 point) to definitely ger number of students thought they belonged to middle
apply (5 points). Total score of this scale was 20–100 class, both in junior high school students (68.9%) and
Li et al. BMC Pediatrics (2020) 20:271 Page 5 of 10

senior high school students (71.0%). Nearly two thirds of education level and lower prestige occupation status of
adolescents reported that their parents had a medium those parents reported lower quality child-parent, peer,
level of education, while nearly half reported that their student-teacher relationship than adolescents with
parents had a medium prestige occupation. The average higher family SES. (Table 2).
score (SD) of family, peer and student-teacher relation- Table 3 presented the results of hierarchical multiple
ship were 18.18 (4.20), 64.90 (13.38) and 76.00 (12.43), linear regression. The Model 1 of child-parent, peer, and
respectively. (Table 1). student-teacher relationships (adjusted R2 was 0.025,
Moreover, all five indicators of SES were weak posi- 0.041, 0.057, respectively) indicated that girls, urban
tively correlated with the quality of social relations (r household registration, intact family were significantly
ranged from 0.036 to 0.189, all p < 0.001), except that associated with higher quality social relationships. Add-
maternal education level was not correlated with the itionally, non-only child had higher quality of parent-
conflict dimension of parent-child relationship. To be child relationship, and older adolescents had lower qual-
specific, adolescents with lower subjective SES, lower ity of peer and student-teacher relationship.

Table 1 Demographic characteristics of the adolescents in East China (n = 6902)

2
Characteristics Junior high school (n = 4208) Senior high school (n = 2694) X p
Gender
Boys 2029 (48.2) 1326 (49.2) 0.661 0.416
Girls 2179 (51.8) 1368 (50.8)
Household registration
Urban 1886 (44.8) 992 (36.8) 43.205 < 0.001
Rural 2322 (55.2) 1702 (63.2)
Family structure
Intact family 3516 (83.6) 2315 (85.9) 7.076 0.008
Non-intact family 692 (16.4) 379 (14.1)
Family size
Non-only child 1997 (47.5) 1535 (57.0) 59.592 < 0.001
Only child 2211 (52.5) 1159 (43.0)
Subjective SES
Lower class 605 (14.4) 517 (19.2) 81.416 < 0.001
Middle class 2899 (68.9) 1914 (71.0)
Upper class 704 (16.7) 263 (9.8)
Paternal education
Low 67 (1.5) 75 (2.8) 17.935 < 0.001
Medium 2607 (62.0) 1728 (64.1)
High 1534 (36.5) 891 (33.1)
Maternal education
Low 243 (5.7) 205 (7.6) 18.848 < 0.001
Medium 2742 (65.2) 1812 (67.3)
High 1223 (29.1) 677 (25.1)
Paternal occupation
Low 68 (1.6) 46 (1.7) 3.986 0.136
Medium 2299 (54.6) 1535 (57.0)
High 1841 (43.8) 1113 (41.3)
Maternal occupation
Low 612 (14.5) 359 (13.3) 20.774 < 0.001
Medium 2061 (49.0) 1470 (54.6)
High 1535 (36.5) 865 (32.1)
Li et al. BMC Pediatrics (2020) 20:271 Page 6 of 10

Table 2 Spearman’s correlation coefficients between SES indicators and different domains of social relationships among adolescents
SES Parent-child relationship Peer relationship Student-teacher relationship
Cohesion Expressiveness Conflict Social Communicative Interpersonal Intimacy Support Satisfaction Conflict
emotion interaction concordance
Subjective SES 0.037** 0.054** 0.048** 0.120** 0.152** 0.168** 0.087** 0.084** 0.092** 0.036**
Paternal education level 0.110** 0.080** 0.049** 0.158** 0.153** 0.189** 0.119** 0.094** 0.082** 0.059**
Maternal education level 0.062** 0.060** 0.023 0.165** 0.156** 0.188** 0.134** 0.087** 0.079** 0.089**
Paternal occupation status 0.112** 0.067** 0.061** 0.142** 0.137** 0.165** 0.095** 0.083** 0.063** 0.049**
Maternal occupation status 0.056** 0.041** 0.049** 0.091** 0.083** 0.117** 0.090** 0.056** 0.052** 0.047**
** Correlations were significant at the 0.01 level (2-tailed)

Model 2 included the five SES indicators and the con- [40]. There are strong relationships between subjective
trol variables (i.e., age, gender, household registration, SES and health outcomes, which can be explained by
family structure, family size). In parent-child relationship several reasons [6]. First of all, subjective SES may reflect
model (adjusted R2 = 0.035), the two single indicators of a relative social status of one person in social class, mod-
SES were statistically significant. Standardized regression erate the relationship between income inequality and
coefficients indicated paternal education level (β = 0.08) population health, rather than demonstrate the absolute
and paternal occupation status (β = 0.07) were the pre- status of one [41]. Secondly, subjective SES may be a
dictors of parent-child relationship. Peer relationship more accurate measurement of social status that taken
model (adjusted R2 = 0.081) revealed that higher subject- past and future prospects into account, and making
ive SES, higher level of parents’ education, and higher more nuanced judgments for objective indicators, which
prestige paternal occupation status meant the adoles- can represent the cognitive average of various socioeco-
cents had a better relationship with peers. The corre- nomic status indicators [5]. Lastly, a reciprocal relation-
sponding effect sizes were small and slightly stronger for ship may exist between subjective SES and health [3].
subjective SES (β = 0.10) than other three SES indicators On the other hand, our analysis also provided evidence
(all β = 0.08). In student-teacher relationship model (ad- that subjective SES was a relatively stronger predictor of
justed R2 = 0.072), it was interesting that subjective SES, peer relationship than objective SES. Adolescents are in
parents’ education level and fathers’ occupation status a special period when they may develop a sense of social
were still significant, but the maternal education level status during the transition between childhood domi-
only had a slightly stronger correlation (β = 0.07) than nated by family status and adulthood dominated by self-
other indicators. determination. The ecological environment constructed
in the process of human development is a dynamic sys-
Discussion tem. So adolescents are more involved in the social en-
Greater attention has been paid to determinants of the vironment outside the family, especially the interaction
health and well-being of adolescents, including SES and with peers. And a number of studies confirmed that the
the social environment of family and school [39]. In this content of friendships changes as children enter adoles-
school-based cross-sectional study, we found that ado- cence [42]. At the early stage of adolescence, individuals
lescents with low SES had lower quality of social rela- have a superficial understanding of friendship and pay
tionships. We also noticed that subjective SES was a attention to common activities. In the middle of puberty,
relatively stronger predictor for peer relationship than there are more emphasis on mutual emotional depend-
objective SES, while objective SES could better predict ence between peers, especially on loyalty, trustworthi-
parent-child relationship and student-teacher relation- ness and respect. Older adolescents generally believe
ship among adolescents. These data have provided valu- that friends need to understand and support each other,
able information for improving social relationships and which involves deep psychological consistency in per-
promoting health among adolescents. sonality. Previous studies have shown that subjective
In this study, adolescents’ self-reported subjective SES SES in adolescence may reflect the influence of objective
and objective SES indicators (parental occupation and SES and modern consumer culture (emphasis on
education) were used to observe their social relationship. property, brand and/or appearance) [43]. Addition-
And as expected, we found that adolescents with higher ally, adolescents with lower subjective SES are more
subjective SES had better relationship with their peers. likely to experience higher stress from different areas
These results extend previous studies that positive peer including adverse social relationships [44]. Adoles-
and student-teacher relationships are more likely to be cents with higher SES may have more sense of su-
established by adolescents with higher subjective SES periority and self-identity than other peers. They are
Li et al. BMC Pediatrics (2020) 20:271 Page 7 of 10

Table 3 Hierarchical multiple regression of social relationships with SES variables among adolescents
Variables Model 1 Model 2
B 95% (CI) p β B 95% (CI) p β
Parent-child relationship
Age (Years) −0.02 − 0.08, 0.04 0.52 −0.01 − 0.01 − 0.07, 0.05 0.70 − 0.01
Gender (Boy vs. Girls) 0.73 0.53, 0.92 < 0.001 0.09 0.71 0.51, 0.91 < 0.001 0.08
Household registration (Urban vs. Rural) −0.23 − 0.44, − 0.03 0.03 − 0.03 0.07 − 0.15, 0.30 0.53 0.01
Family structure (Intact vs. Non-intact) −1.38 −1.65, −1.11 < 0.001 − 0.12 − 1.28 − 1.55, − 1.01 < 0.001 − 0.11
Family size (Non-only child vs. only child) − 0.52 − 0.72, − 0.31 < 0.001 − 0.06 − 0.49 − 0.70, − 0.28 < 0.001 −0.06
Subjective SES 0.18 −0.01, 0.37 0.06 0.02
Paternal education level 0.65 0.42, 0.89 < 0.001 0.08
Maternal education level −0.20 −0.42, 0.03 0.09 −0.03
Paternal occupation status 0.54 0.32, 0.77 < 0.001 0.07
Maternal occupation status −0.10 −0.27, 0.07 0.25 −0.02
Adjusted R2 0.025 0.035
ΔR2 0.026 0.011
Peer relationship
Age (Years) −0.42 −0.6, −0.23 < 0.001 −0.05 − 0.29 −0.48, − 0.11 < 0.001 −0.04
Gender (Boy vs. Girls) 1.95 1.33, 2.58 < 0.001 0.07 1.71 1.10, 2.33 < 0.001 0.06
Household registration (Urban vs. Rural) −3.40 −4.05, −2.75 < 0.001 −0.13 −0.94 −1.64, − 0.24 0.01 − 0.04
Family structure (Intact vs. Non-intact) −4.43 −5.29, − 3.57 < 0.001 − 0.12 − 3.73 −4.58, − 2.89 < 0.001 −0.10
Family size (Non-only child vs. only child) −0.46 − 1.11, 0.19 0.17 −0.02 0.09 −0.56, 0.74 0.79 0.003
Subjective SES 2.45 1.86, 3.03 < 0.001 0.10
Paternal education level 2.04 1.30, 2.77 < 0.001 0.08
Maternal education level 1.92 1.21, 2.62 < 0.001 0.08
Paternal occupation status 2.07 1.38, 2.77 < 0.001 0.08
Maternal occupation status −0.12 −0.65, 0.41 0.67 −0.01
Adjusted R2 0.041 0.081
ΔR2 0.041 0.041
Student-teacher relationship
Age (Years) −1.17 −1.34, −1.00 < 0.001 −0.16 −1.11 −1.28, − 0.93 < 0.001 −0.15
Gender (Boy vs. Girls) 2.29 1.71, 2.87 < 0.001 0.09 2.12 1.54, 2.7 < 0.001 0.09
Household registration (Urban vs. Rural) −1.44 −2.04, −0.84 < 0.001 − 0.06 0.06 − 0.6, 0.71 0.87 0.002
Family structure (Intact vs. Non-intact) −4.80 −5.58, −4.01 < 0.001 −0.14 −4.37 −5.16, −3.58 < 0.001 − 0.13
Family size (Non-only child vs. only child) −0.05 − 0.65, 0.55 0.88 − 0.002 0.36 − 0.25, 0.96 0.25 0.01
Subjective SES 0.90 0.36, 1.44 < 0.001 0.04
Paternal education level 1.15 0.47, 1.84 < 0.001 0.05
Maternal education level 1.52 0.86, 2.18 < 0.001 0.07
Paternal occupation status 1.13 0.48, 1.77 < 0.001 0.05
Maternal occupation status 0.20 −0.29, 0.7 0.42 0.01
2
Adjusted R 0.057 0.072
ΔR2 0.057 0.016

also more attractive and mature, which makes them Focusing on objective SES, our result revealed that the
easier to be accepted by peers in interpersonal com- education of father and mother was the main predictor
munication, thus satisfying their needs for peers at of parent-child relationship and student-teacher rela-
different times [45]. tionship, respectively. This is consistent with a number
Li et al. BMC Pediatrics (2020) 20:271 Page 8 of 10

of studies that have shown a relevance between parental social desirability bias. While there are some limitations
higher education and various positive health outcomes in the current study. First of all, our analysis was based
[46]. Furthermore, the most widely used measure stand- on cross-sectional data, which limited our ability to con-
ard of SES among adolescents is parental education level fidently infer the direction of causality. Secondly, income
[47]. The advantage of the index is that it is easy to as one of the main indicators of socioeconomic status
measure and it can strongly predict employment and was not included in this study, because it was generally
income levels [48], which is considered to be the most considered as sensitive information in China and the au-
effective predictor of adolescent health [2]. thenticity of the obtained data couldn’t be guaranteed.
Knowledge, social status and available resources could But researchers are encouraged to include this indicator,
be all reflected in education. And the higher of parents’ as it is relatively independent influence [51]. Thirdly, the
education, the more conducive to develop adolescents’ study only analyzed the socioeconomic status of adoles-
interpersonal relationships [49]. Family is the first and cents’ family, while ignoring their school SES. Finally,
most important factor that influences children’s the representativeness of the results may be limited, as
socialization. As important persons in children’s life, only middle school students’ data were collected in this
parents have close emotional contacts with children, and study and the same number of schools were selected
the most frequent participation and management in from three cities without the estimated weight adjust-
their children’s social life. Therefore, parents have very ment. Nevertheless, it had substantive, practical and
important influences on all aspects of children’s life, in- methodological implications, which increased small body
cluding interpersonal relationship [50]. It is noticed that of work on socioeconomic status and social relations.
some behavioral influences posed on children’s develop-
ment, such as lifestyle choices, parenting styles, know- Conclusion
ledge and skills are linked to parental education [51]. Substantively, we found the inequity existed in social re-
Highly educated parents have a better understanding of lationships of adolescents from East China. Adolescents
child development and are able to choose a more appro- with lower SES had poorer social relationships relative
priate parenting type and practice [52]. Parents play dif- to higher SES. And subjective SES, paternal education
ferent roles in the family, under the influence of Chinese and maternal education were the main predictors of peer
traditional social culture, mothers often spend a lot of relationship, parent-child relationship and student-
time in the family and have more responsibility for tak- teacher relationship, respectively. An understanding of
ing care and educating adolescents. While fathers pay the effects of socioeconomic on social relationships will
more attention to the development of the career and prompt public health experts and policy makers to iden-
provide solid economic foundation for the family, who tify, intervene, and eventually alleviate the root causes of
tend to lack of effective and in-depth communication adolescents’ health-associated problems. As such, pre-
with adolescents [53]. In addition, during the transition ventative programs and services for adolescents with low
from childhood to adulthood, adolescents become more parental SES as well as low subjective SES should be
self-awareness and are more involved in social relations provided. These programs should include social support
outside the family along with the rapid, novel and unex- and education to raise awareness of the problems faced
pected changes occurred in the physical, psychological, by adolescents and their families and to address them so
and social development [15]. Therefore, in the process as to improve their outcomes and avoid the negative ef-
of raising adolescents by parents, mothers with high fects of low SES [56].
education level can better guide adolescents to establish
positive social relationship, including student-teacher re- Abbreviations
SES: Socioeconomic status; FES: Family Environment Scale; ISCO: International
lations [54]. Highly educated fathers are more likely to Standard Classification of Occupations; ISCED: International Standard
maintain a good parent-child relationship, for they are Classification of Education
more aware of the problems and adverse effects, and can
provide measures to improve the situation [55]. Acknowledgements
Not applicable.
This school-based study may lead to a deeper and
more diverse understanding of the relationship between Authors’ contributions
adolescent social relationships and socioeconomic fac- YX designed and directed the study; JL collected and analyzed the data, JL
and JW drafted the manuscript; literature searched by RXJ, JYL and SQ; JL,
tors. Our research focuses on the multi-dimensional JHL and YQW conducted the field work. All authors participated in the
family socioeconomic status in attempt to capture the interpretation of data and the preparation of the manuscript as well as read
social status of family differentially and extensively. Add- and approved its final version. YX took responsibility for the integrity of the
data and the accuracy of the data analysis.
itionally, all measurements were self-reported, except
that students were asked to fill out questionnaires an- Funding
onymously without the presence of teachers to reduce Not applicable.
Li et al. BMC Pediatrics (2020) 20:271 Page 9 of 10

Availability of data and materials 15. U B. The ecology of human development. Cambrigde: Harvard University
Datasets used and analyzed during this study are available from the Press; 1979.
corresponding author on reasonable request. 16. Vanassche SSAK, Matthijs K, Swicegood G. The effects of family type, family
relationships and parental role models on delinquency and alcohol use
Ethics approval and consent to participate among Flemish adolescents. J Child Fam Stud. 2013;23(1):128–43.
This study obtained ethical approval from the Research Center for Health 17. Hartup WW. The company they keep: friendships and their developmental
Care Development of Soochow University. All participants and their parent significance. Child Dev. 1996;67(1):1–13.
or guardians were informed about the study and provided verbal informed 18. Sentse M, Lindenberg S, Omvlee A, Ormel J, Veenstra R. Rejection and
consent before the study was conducted. Verbal informed consent was acceptance across contexts: parents and peers as risks and buffers for early
approved by the ethics committee. adolescent psychopathology. The TRAILS study. J Abnorm Child Psychol.
2010;38(1):119–30.
19. Longobardi C, Prino LE, Marengo D, Settanni M. Student-teacher
Consent for publication relationships as a protective factor for school adjustment during the
Not applicable. transition from middle to high school. Front Psychol. 2016;7:1988.
20. Fowler LTS, Banks T, Anhalt K, Der HH, Kalis T. The association between
Competing interests externalizing behavior problems, teacher-student relationship quality, and
The authors declare that they have no conflicts of interests. academic performance in Young urban learners. Behav Disord. 2008;33(3):
167–83.
Author details 21. Hamre BK, Pianta RC. Early teacher-child relationships and the trajectory of
1 children's school outcomes through eighth grade. Child Dev. 2001;72(2):
Department of Child and Adolescent Health, School of Public Health,
Medical College of Soochow University, No.199 Ren Ai Road, Suzhou, Jiangsu 625–38.
215123, People’s Republic of China. 2Business School, Xi’an International 22. Umberson DMJK. Social relationships and health: A flashpoint for health
Studies University, Xi’an 710128, People’s Republic of China. 3Department of policy. J Health Soc Behav. 2010;51(Supp l):54–66.
Social medicine, School of Public Health, Soochow University, Suzhou, 23. Zhang M, Wang Z. Mental health state of middle or high school students.
Jiangsu 215123, People’s Republic of China. Chin Ment Health J. 2001;15(4):226–7.
24. Braveman PA, Cubbin C, Egerter S, Chideya S, Marchi KS, Metzler M, Posner
Received: 7 November 2019 Accepted: 27 May 2020 S. Socioeconomic status in health research: one size does not fit all. Jama.
2005;294(22):2879–88.
25. [Link]. Author@ce: The Psychology of Survey
Response; 2000.
References
1. Cirino PT, Chin CE, Sevcik RA, Wolf M, Lovett M, Morris RD. Measuring 26. OECD: ISCED 2011 Operational Manual Guidelines for Classifying National
socioeconomic status: reliability and preliminary validity for different Education PrograMmes and Related Qualifications 2015.
approaches. Assessment. 2002;9(2):145–55. 27. Lien N, Friestad C, Klepp KI. Adolescents’ proxy reports of parents’
2. Gakidou E, Cowling K, Lozano R, Murray CJ. Increased educational socioeconomic status: How valid are they? J Epidemiol Commun Health.
attainment and its effect on child mortality in 175 countries between 1970 2001;55(10):731–7.
And 2009: a systematic analysis. Lancet (London, England). 2010;376(9745): 28. Office IL. International standard classification of occupations: ISCO-08 (Vol.1)
959–74. structure, group definitions and correspondence tables. Geneva:
3. Garbarski D. Perceived social position and health: is there a reciprocal International Labour Office; 2012.
relationship? Soc Sci Med. 2010;70(5):692–9. 29. Goodman E, Adler NE, Kawachi I, Frazier AL, Huang B, Colditz GA.
4. Adler NE, Epel ES, Castellazzo G, Ickovics JR. Relationship of subjective and Adolescents' perceptions of social status: development and evaluation of a
objective social status with psychological and physiological functioning: new indicator. Pediatrics. 2001;108(2):E31.
preliminary data in healthy white women. Health Psychol. 2000;19(6):586– 30. MRH HCJ. The quality of social support: Measures of family and work
92. relationships. Br J Clin Psychol. 1983;22(3):157–62.
5. Singh-Manoux A, Adler NE, Marmot MG. Subjective social status: its 31. Xue L, Zhu X, Bai M, et al. The Chinese Version of Family Environment Scale
determinants and its association with measures of ill-health in the Whitehall Symptoms Questionnaire in Adolescent Students: Its Reliability and Validity.
II study. Soc Sci Med. 2003;56(6):1321–33. Chin J Health Psychol. 2014;22(6):881–3.
6. Singh-Manoux A, Marmot MG, Adler NE. Does subjective social status 32. Wei Y. Influence of school factors on self-esteem development among
predict health and change in health status better than objective status? adolescents and children. Psychol Dev Educ. 1998;02:12–6.
Psychosom Med. 2005;67(6):855–61. 33. Pan Y. Research on the status of peer relationship satisfaction of hearing-
7. Mackenbach JP. Health inequalities: lifecourse approaches. BMJ Clin Res. impaired middle school students. Foundations Educ Res. 2015;13(7):73–5.
2003;60(3):387–90. 34. Pianta RC. Student-teacher relationship scale: professional manual:
8. Kramer MS, Seguin L, Lydon J, Goulet L. Socio-economic disparities in psychological assessment resources; 2001.
pregnancy outcome: why do the poor fare so poorly? Paediatr Perinat 35. Qu Z. The characteristics of classroom environment in primary and
Epidemiol. 2000;14(3):194–210. secondary schools and its relationship with students’ school adaptation.
9. Adler NE, Ostrove JM. Socioeconomic status and health: what we know and Beijing: Beijing Normal University; 2002.
what we don't. Ann N Y Acad Sci. 1999;896:3–15. 36. Yan J. The relationship between interpersonal closeness, mental health and
10. Viner RM, Ozer EM, Denny S, Marmot M, Resnick M, Fatusi A, Currie C. academic performance of junior high school students and its intervention.
Adolescence and the social determinants of health. Lancet (London, Engl). Shanghai: Shanghai Normal University; 2018.
2012;379(9826):1641–52. 37. Cohen J. In: 2nd ed, editor. Statistical power analysis for the behavioral
11. Chen E, Paterson LQ. Neighborhood, family, and subjective socioeconomic sciences. Hillsdale: Lawrence Erlbaum Associates; 1988.
status: how do they relate to adolescent health? Health psychology : official 38. Nagelkerke NJD. A note on a general definition of the coefficient of
journal of the division of Health Psychology. Am Psychol Assoc. 2006;25(6): determination. Biometrika. 1991;78(3):691–2.
704–14. 39. Morrow V. Conceptualising social capital in relation to the well-being of
12. Lowry R, Kann L, Collins JL, Kolbe LJ. The effect of socioeconomic status on children and young people: a critical review. Soc Rev. 2010;47(4):744–65.
chronic disease risk behaviors among US adolescents. JAMA. 1996;276(10): 40. Ye Z, Wen M, Wang W, Lin D. Subjective family socio-economic status,
792–7. school social capital, and positive youth development among young
13. West P. Health inequalities in the early years: is there equalisation in youth? adolescents in China: a multiple mediation model. Int J Psychol. 2020;55(2):
Soc Sci Med. 1997;44(6):833–58. 173–81. [Link]
14. Goodman E. The role of socioeconomic status gradients in explaining 41. Wilkinson RG. Health, hierarchy, and social anxiety. Ann N Y Acad Sci. 1999;
differences in US adolescents' health. Am J Public Health. 1999;89(10): 896:48–63.
1522–8. 42. Brown BB. Adolescents’ relationships with peers; 2013.
Li et al. BMC Pediatrics (2020) 20:271 Page 10 of 10

43. Karvonen S, Rahkonen O. Subjective social status and health in young

people. Soc Health Illness. 2011;33(3):372–83.
44. Bradley RH, Corwyn RF. Socioeconomic status and child development. Annu
Rev Psychol. 2002;53:371–99.
45. Sweeting H, West P, Young R, Kelly S. Dimensions of adolescent subjective
social status within the school community: description and correlates. J
Adolesc. 2011;34(3):493–504.
46. Wilkinson RG, Pickett KE. The problems of relative deprivation: why some
societies do better than others. Soc Sci Med. 2007;65(9):1965–78.
47. Svedberg P, Nygren JM, Staland-Nyman C, Nyholm M. The validity of
socioeconomic status measures among adolescents based on self-reported
information about parents occupations, FAS and perceived SES; implication
for health related quality of life studies. BMC Med Res Methodol. 2016;16:48.
48. Galobardes B, Lynch J, Smith GD. Measuring socioeconomic position in
health research. Br Med Bull. 2007;81-82:21–37.
49. Xie B. Education degree of parents, parenting style and children’s
interpersonal skills. Hunan: Xiangtan University; 2019.
50. Chao RK. Beyond parental control and authoritarian parenting style:
understanding Chinese parenting through the cultural notion of training.
Child Dev. 1994;65(4):1111–9.
51. Conger RD, Donnellan MB. An interactionist perspective on the
socioeconomic context of human development. Annu Rev Psychol. 2007;58:
175–99.
52. Mohammadi M, Zarafshan H. Family function, parenting style and broader
autism phenotype as predicting factors of psychological adjustment in
typically developing siblings of children with autism Spectrum disorders.
Iran J Psychiatry. 2014;9(2):55–63.
53. Yan Xulei YP. A new theory of family education. Beijing: Peking University
Press; 2012.
54. Fox RA, Platz DL, Bentley KS. Maternal factors related to parenting practices,
developmental expectations, and perceptions of child behavior problems. J
Genet Psychol. 1995;156(4):431–41.
55. Paquette D. Theorizing the father-child relationship: mechanisms and
developmental outcomes. Hum Dev. 2004;47(4):193–219.
56. Huang L, Stroul B, Friedman R, et al. Transforming mental health care for
children and their families. Am Psychol. 2005;6(60):615–27.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.