Skip to main content

Co-occurrence subgroups of child sexual abuse, health risk behaviors and their associations among secondary school students in China

Abstract

Background

Little is known on the co-occurrence and heterogeneity of child sexual abuse (CSA) or health risk behavior (HRB) prevalence nor the associations among the victims.

Objectives

To detect the prevalence and subgroups of adolescents reporting CSAs or HRBs, and to examine the association between the subgroups.

Methods

Participants were secondary school students in a national survey in China (N = 8746). Self-reported CSA and HRB experiences were collected through a computer assisted questionnaire. Prevalence and confidence intervals were calculated. Multigroup latent class analysis (LCA) was used to examine latent subgroups of CSA and HRB. Dual latent class regression analysis was used to examine the association between CSA and HRB classes.

Results

A total of 8746 students participated in our study. The prevalence of having ever experienced any of the reported seven CSA items was 12.9%. The preferred LCA model consisted of a three-class CSA latent variable, i.e. “Low CSAs”(95.7% of the total respondents), “Verbal or exhibitionism CSAs”(3.3%), and “high multiple CSAs” (1.1%); and a three-class HRB latent variable, i.e. “Low HRBs”(70.5%), “externalizing HRBs” (20.7%), and “internalizing HRBs” (8.7%). Students in the “Verbal or exhibitionism CSAs” or “high multiple CSAs” classes had higher probabilities of being in “externalizing HRBs” or “internalizing HRBs” classes. The probabilities were higher in “high multiple CSAs” class(male externalizing OR 4.05, 95%CI 1.71–9.57; internalizing OR 11.77, 95%CI 4.76–29.13; female externalizing OR 4.97, 95%CI 1.99–12.44; internalizing OR 9.87, 95%CI 3.71–26.25) than those in “Verbal or exhibitionism CSA”(male externalizing OR 2.51, 95%CI 1.50–4.20; internalizing OR 3.08, 95%CI 1.48–6.40; female externalizing OR 2.53, 95%CI 1.63–3.95; internalizing OR 6.05, 95%CI 3.73–9.80).

Conclusions

Prevalence of CSA items varies. Non-contact CSAs are the most common forms of child sexual abuse among Chinese school students. There are different latent class co-occurrence patterns of CSA items or HRB items among the respondents. CSA experiences are in association with HRB experiences and the associations between latent classes are dose-responded. Multi-victimization has more significantly negative effects. The results could help identify high-risk subgroups and promote more nuanced interventions addressing adverse experiences and risk behaviors among at-risk adolescents.

Peer Review reports

Background

Child Sexual Abuse (CSA) means the involvement of a child in sexual activity that he or she does not fully comprehend, is unable to give informed consent to, or for which the child is not developmentally prepared and cannot give consent, or that violates the laws or social taboos of society [1]. CSA has spread throughout countries, continents, and socioeconomic classes in recent decades [2] and attracted increasing attention from the public, the media, and the academia. WHO [3] estimated in 2006 that 150 million girls and boys under the age of 18 have experienced CSA. Some meta-analyses estimated the average global prevalence of CSA to be around 11.8%, although it varies from 2 to 62% among different studies as well as types of CSA [4,5,6].

CSA is regarded as a type of Adverse Childhood Experiences. Studies, particularly among youth, have shown CSA victims are more likely to engage in behaviors that can place them at risk for negative outcomes, which are in general referred to as youth Health Risk Behavior (HRB), such as alcoholism, drug abuse, anxiety, violence, and suicidal ideation [7, 8], and has long been concerned with short-term and long-term negative sequelae [9,10,11,12,13,14,15,16]. Even a single form of abuse can be associated with risk in different HRB domains [17,18,19]. CSA is however a multi-indicator measurement, which indicates that experiencing CSA often means experiencing more than one form of CSA [20, 21] i.e. multiple victimization [22, 23]. While the consequence of CSA may not be measurably apparent for some youth, other youth report HRBs}{suicidal ideation [24, 25]. Different forms of abuse items may overlap and interact with each other, and the combination of different forms may modify specific social and psychological consequences [23, 26, 27]. CSA’s covert nature may prevent victims from expressing themselves and seeking help. Without considering the nature of co-occurrence, the estimation of the association between CSA and HRB may be biased [28]. The lack of research on multi-victimization hinders our understanding of the impact of multiple CSA victimization [29]. Insights into different subgroups of CSA and the association patterns with HRBs could help service providers and decision makers identify adolescents at risk that need help, and design specific inventions according to victims’ characteristics and demands [30,31,32].

China has the largest youth population in the world, with more than 200 million students in primary and secondary schools, accounting for 20% of the national population in total. CSA has attracted public attention in China, while researches in this field are still insufficient, with scarcely any on multiple victimization. Existing Chinese researches on the association between child victimization and health behaviors often focused on single form of the adolescent adverse experiences from small samples or limited study sites [33,34,35]. It is thought that the cultures and traditions of Asian societies may lead to different CSA patterns and their associations with behaviors, but there is little research on CSA patterns [36] in China. There is need to examine the heterogeneity and patterns of multi-victimization in teens to help optimize prevention and intervention programs on a national scale.

Methods

Study aims

This study aimed to investigate the prevalence of CSA and HRB along with the latent co-occurrence patterns, and to test the hypothesis that the subtypes of CSA are associated with HRB sub-types.

Survey design, sites and population

This was a cross-sectional school-based study. The data for the analysis came from a national reproductive and health survey of secondary school students in China. Given China’s population diversity, multistage sampling was applied to recruit study participants. Seven provinces/autonomous regions (Shandong, Guangxi, Hebei, Heilongjiang, Sichuan, Shaanxi, Inner Mongolia) representing geographic or social variation were selected. Then, one city/town with a moderate level of economic development was selected in each province/autonomous region. In each city/town, one urban area and one rural area were selected, after which one junior secondary school and one senior secondary school were selected in each area. Two classes in each grade (including all three grades in junior secondary school and all three grades in senior secondary school) were invited to participate in the survey using cluster random sampling.

Survey implementation and data collection

This survey was conducted between year 2014 and 2015. Data collection was through a questionnaire with a computer-assisted self-interviewing (CASI) approach. The content of the questionnaire related to this analysis included a set of designed questions on experiences of CSA items, experiences of HRB items, and some covariates. During the survey day, each student took a separate seat in the school’s computer room and completed the electronic questionnaire. The questionnaire was anonymous, and the subjects could not obtain information of others even using the same computer afterwards. Field investigators assisted the students by monitoring or providing explanations when necessary to avoid ambiguity, however, they could not see or interfere with the respondent’s answers. If any respondent feels uncomfortable with any question, he or she can skip part or all of the content. The data generated after the investigation were uniformly stored and processed by the researchers after deprivacy.

CSA scale designing and testing

CSA refers to reluctant sexually related experiences encountered by the participated students aged 10-18 years. Sexual abuse experiences from peers were excluded. Due to the length of the investigation, we modified the ISPCAN Child Abuse Screening Tool (ICAST-R) [37] and Chen’s instrument [38] to obtain a set of seven-item self-reported CSA scale before the age of 18. Chen’s 12-item CSA questionnaire is the most widely used CSA scale in Chinese CSA study [39]. Compared with Chen’s instrument, our CSA measurements made these following adjustments: adding the question “Having been told dirty jokes or shown pornographic pictures, publications or supplies, etc” as one form of non-physical contact CSA, deleting the question “Tried to sexually arouse the child” and “Made child arouse them and touch their body in a sexual way” because they were hard to understand for children, deleting “Tried to have anal intercourse with child” and “Had anal intercourse with child” due to its abstruseness and the rare prevalence reported, combining “Exposed their genitals to the child” and “Masturbated in front of the child” into one question” Having seen someone exposing his/her genitals or masturbating in front of you”, combining “Touched child’s genitals with their mouth” and “Made child touch their genitals with child’s mouth” into one question” Having had someone touching your privates/breasts or forcing you to touch his/her privates/breasts”. Our seven-item two-category (yes/no) self-reported CSA scale has a Cronbach’s alpha 0.65 [37,38,39].

Measures

Child sexual abuse (CSA): The study includes a set of seven-item two-category (yes/no) self-reported CSA scale (Table 1).

Table 1 Definitions of CSA and HRB items

Health risk behaviors (HRB): The study includes several risk behaviors that cover different behavioral dimensions. These questions were transformed into a set of two-category items (yes / no) (Table 1). The HRB scale has a Cronbach’s alpha 0.69.

Covariates

Covariates in this analysis were demographic characteristics including the respondents’ residential area, age, school grade, whether to live on campus, peer relationship, overall feeling of school, academic performance, the respondents’ free time lifestyle such as reading/playing video games/chatting/regular physical exercise/internet surfing, and the respondents family characteristics such as number of siblings, family economic condition, overall feeling of family, parent’s education, relationship with parents, parents disciplinary (Table 2).

Table 2 Participant demographic characteristics

Ethics

The research has been reviewed and approved by the local ethics committee. Before the survey was conducted, the purpose and implementation of the study were explained to the students, teaching staff, and parents in the parents meeting whose children were on the invitation list of the sample classes. Anonymity and confidentiality were guaranteed and the rights to refuse or terminate participation in the investigation were understood. Students and parents who were willing to accept the invitation to the study were asked and signed a written consent form.

Power calculation

Based on the sampling framework, about 50 students were planned to investigate of each sex and each age group from each site and the expected sample size was 8400. All students attended to school in the sample classes on the survey day were invited to our survey. Finally, we yielded a total sample of 8746 students. In this study, the associations between CSAs and HRBs were analyzed. According to the literature [32], the OR between CSAs and HRBs, such as depression, suicidal ideation and substance abuse, etc., was between 1.2 and 2.9. Considering 2 as the estimated OR value, we conducted the power estimation based on the total sample size of 8746 cases, and then the statistical power(1-β) could reach 82.7%.

Analytic approach

Descriptive statistics were conducted to examine the prevalence of CSA items and HRB items. Categorical data were reported in count and percentage. Then a multi-group latent class analysis (LCA) was adopted to detect profiles of CSA and profiles of HRB across genders to explain the relation between several categorical manifest variables (indicators) by one or more underlying latent categories (classes). Model fit statistics such as the Bayesian information criterion (BIC), the Akaike information criterion (AIC), and the interpretability were considered when evaluating model fit. Once the preferred LCA models were conducted, we got two latent categorical variables (CSA profile and HRB profile). We determined the most likely latent class membership using the highest posterior class probability for each individual. To explore the association of CSA profile and HRB profile, dual LCA regression stratified by gender was used to explore the association of the two latent categorical variables. Covariates can be incorporated in the dual LCA regression model using a logistic link function. Missing data were replaced using full information maximum likelihood (FIML) estimation, which is the default in Mplus. Sawtooth Software was used to classify data from the questionnaire. Data was checked and cleaned site by site, and then merged and exported into Stata format via Stata version 15.1. Descriptive analysis was conducted in R version 3.6.1. Latent class analysis and the dual LCA regression were completed using Mplus version 7.4 [40].

Results

Participant demographics

According to the list of survey respondents, 8910 students were consented to take part in the survey and supposed to be at school on the date of survey. However, 164 were absent due to sickness or personal affairs on the actual survey date. As a result, the total number of respondents included in the survey were 8746, with a response rate of 98.2%. Ages of the participants ranged from 12 to 18 years, with a mean of 15.37 ± 1.95. The number of male students participated in the survey was 4339(49.6%), and that of female students was 4407(50.4%). Proportions of other characteristics were shown in Table 2.

Prevalence of CSA and HRB items

The prevalence of the CSA indicators is shown in the upper half of Table 3. In general, a total of 1131 respondents reported to have had at least one CSA experience, with an overall proportion of 12.9%. Among all respondents, 527 (6.0%) had experienced non-physical contact CSA, while 357 (4.1%) had experienced physical-contact CSA with 247(2.8%) having had experienced attempted intercourse or forced intercourse. Between the genders, non-physical contact CSA and physical-contact CSA were more commonly experienced while attempted intercourse or forced intercourse were less seen in males. The prevalence of CSA stratified by covariates such as individual, familial or school factors could be seen in Additional Table 1.

Table 3 Prevalence of CSA and HRB by gender and type of abuse (N (%))

The prevalence of the HRB indicators are shown in the lower part of Table 3. There were 39.4% that had at least one risk behavior. For males, there was a higher prevalence of “sex-related behavior”, “smoking”, “drinking alcohol”, “gambling”, “skipping classes “, “fighting” and “running away from home”. For females, there was a higher prevalence of having a suicidal ideation, (female 14.1% vs. male 10.3%) or suicidal behavior (female 3.2% vs. male 1.9%).

LCA model selection

Table 4 shows the fit statistics of the latent class analysis models for CSAs and HRBs. To select a suitable latent class model for the seven CSA items, multi-group LCA models were estimated from two to nine latent classes. As can be seen in Table 4, the three-class solution CSA latent class had the lowest BIC value (25,544.357), and the lowest AIC value (25,214.351). As for HRB, the three-class solution has the lowest BIC (71,091.7). Although the AIC value at this point was not the smallest, there was an obvious turning point to show a steep drop before the point and a slow descent after the point. Meanwhile, the three-class model has a clear, meaningful interpretation. The probabilities of classification error of both models were fairly accepted (CSA 0.0369, HRB 0.1314). Therefore, this analysis considered the three-classification model as the most suitable latent class model for CSAs and HRBs.

Table 4 Parameters for LCA model selection

Latent class profiles

As shown in Fig. 1, the three profiles of CSAs were characterized by the following conditional probabilities: In “high multiple CSAs”, there was a high probability of positive response on each CSA item. In “Verbal or exhibitionism CSA”, a higher probability of positive response was seen on the item “Having been told dirty jokes or shown pornographic pictures, publications or supplies. while lower probability on items such as “Having had someone rubbing his/her genitals on you”, “Having had someone touching your genitals or forcing you to make contact with his/her genitals by mouth”, “Having had someone attempting to have sex with you” and “Having had someone forcing to have sex with you”. In “Low CSAs”, there was a low probability of positive response on each topic. Among them, the “Low CSAs” accounted for the largest proportion (boys 96.1%, girls 95.2%). A smaller proportion of respondents was classified as the “Verbal or exhibitionism CSA”(boys 2.6%, girls 3.9%). The number of respondents in the “high multiple CSAs” is the smallest. The proportion of “high multiple CSAs” in boys is higher than that in girls (1.3% vs. 0.9%).

Fig. 1
figure1

Profile probabilities of CSA latent class analysis

As shown in Fig. 2, the three profiles of HRBs were characterized by the following probabilities: The “Low HRBs”, which accounted for the largest proportion (70.5%) in the population, had the lowest probability of positive response on each HRB item. A smaller proportion (20.7%) of the population belonged to “externalizing HRBs” which showed a higher probability of positive response on topics such as “fondling “, “smoking”, “drinking alcohol”, “gambling”, “skipping classes” and “fighting”, while a lower probability of positive response on “depression, despair, or extreme anxiety”, “suicidal ideation”, and “suicidal behavior”. The lowest proportion in the population of this latent class was taken up by “internalizing HRBs”(8.7%), which had a higher probability of positive responses on topics such as “depression, despair, or extreme anxiety”, “suicidal ideation” and “suicidal behavior”, while a lower probability of positive responses on the remaining topics. Compared between genders, girls were more prominent in internalization behaviors, while boys presented a two-peak distribution of both internalization and externalization of the latent conditional probabilities.

Fig. 2
figure2

Profile probability of HRB latent class analysis

Association analysis

In Table 5, results of the dual LCA regression model among boys showed that: considering “Low HRBs” as the baseline, there was a significant association between “externalizing HRBs” and “Verbal or exhibitionism CSA” (OR 2.51, 95% CI 1.50–4.20), and a significant association between the “externalizing HRBs” and “high multiple CSAs” (OR 4.05, 95% CI 1.71–9.57). There was also a significant association between “Internalizing behaviors “ and “Verbal or exhibitionism CSA” (OR 3.08, 95% CI 1.48–6.40); and a significant association between the “ Internalizing behaviors “ and “high multiple CSAs” (OR11.77, 95% CI 4.76–29.13). The association of “Verbal or exhibitionism CSA” and “high multiple CSAs” in girls was of greater strength, which almost doubled that in boys (male OR 3.08, 95% CI 1.48–6.40 vs female OR 6.05, 95% CI 3.73–9.80). After adjusted by variables such as school, family, and personal lifestyle, the above associations still existed, and their directions remained the same, with little difference in strength.

Table 5 Association between CSA classes and HRB classes

Discussion

Across the globe, the self-report prevalence of CSA varies greatly among different studies ranging from 2 to 62% [4,5,6], and the average is estimated to be 12% [4]. According to a meta-analysis that synthesized CSA prevalence of Chinese studies [39], prevalence of CSA showed large variation from 2%(Yen, 2008)to 35.2%(Zhao&Li, 2006)of self-reported at least one form of CSA experiences in China [41]. Different studies have used different CSA items to define CSA, which largely affects the estimated prevalence [42]. Estimates from broad definitions of CSA that includes non-contact CSA yield higher prevalence compared with estimates from narrow definition of CSA that merely include contact CSA (Chen et al., 2010). According to Ji’s review about CSA studies in China, the prevalence of contact CSA varied from 3.3% (Ye, Tao, Fang, Huang, & Sun, 2006) to 14.5% (Chen, Dunne, & Han, 2004), and that of penetrative CSA varied from 0.4% (Chen, Dunne, et al., 2004; Li, 2008) to 2.9% (Chen, Dunne, & Han, 2006) in females. The prevalence of contact CSA varied from 3% (So-kum Tang, 2002) to 15% (Chen, Wang, & Dunne, 2003), and that of penetrative CSA varied from 0 (Li, 2008) to 2.9% (Xia et al., 2009) in males. Our research has produced consistent results. The prevalence of experiencing at least one form of CSAs is 12.9% among the participated students in China. The prevalence of each measured sexual abuse item varies from 1.0 to 4.4%. Non-contact CSAs are the most common forms. Some current studies have shown that the majority of victims were under 14 years old when they encountered CSA [43], and the least age difference between the perpetrator and the victim was 5 years [44]; while in some other studies, the age cut-off point is set as 16 years old [45,46,47] or 18 years old [48, 49]. The discrepancy in study population is also one of the reasons for the difference, as the research objects in Sun’s research are adults [48], while those in Chen’s research [50] are college students in selected areas. Variations of prevalence from different studies are mainly due to different definition of CSA experiences, different constrains to the age of the victim/the perpetrator, different situations in which CSA occurs, and different recruitments of respondents as well. The direct comparison should be done with caution.

Traditional studies recommended CSA to be classified into four types, namely (1) non-contact, (2) genital touching, (3) attempted vaginal or anal insertion, and (4) vaginal or anal penetration. Empirical opinions showed that CSA involving types (2), (3), and (4) are more likely to have significant negative consequences. Other studies suggested to group CSA into two types, one being body-contact CSA and the other being non-contact CSA. However, in real life, victims often encounter more than one type of CSA [51]. Traditional classification methods set empirical cut-off values for classification, dividing children into mutually exclusive categories. One advantage of these methods is that the classification is easy to create and compare. The disadvantage is that they required priori assumptions about the adverse impact, and often took into account only one type of the most serious sexual abuse experience that an individual had experienced. They cannot extract how and to what extent various types of abuse coexist in the target population. Thus, some other studies tried cluster analysis to explore more reasonable CSA classification. For example, a study conducted on 303 teenagers in Denmark resulted in a model of four latent classes, namely, multiple CSA latent class, high language/low contact latent class, high sexual contact latent class and nonvictim latent class [52].

The results of our study are consistent with the literatures highlighting subgroups of CSA by latent class analysis rather than the traditional empirical approach. Compared with the categories obtained according to the traditional classification, the latent class of CSA obtained in our study has taken into account both heterogeneity and co-occurrence. Based on the conditional probability and local independence, this study obtained a three-class latent variable to characterize the latent categorical profiles behind CSA experience as “Low CSAs “, “Verbal or exhibitionism CSA”, and “High Multiple CSAs”. The results of our study and the Danish study found similar latent class subgroups of CSA victims, characterized by high language/low contact latent class, high multiple CSA class and low CSA latent class, which reflected that the LCA method is helpful and stable in detecting latent CSA subgroups. In another study trying using LCA, a sample of 657 young people recruited from high schools and colleges was examined to identify latent class of sexually abusive perpetrators [53]. In this analysis, latent classes of kissing/caressing, attempting sexual intercourse, and completed sexual intercourse were established based on CSA items, which were then found to be related to the abusive behaviors of verbal coercion, material seduction, and violent coercion of the perpetrators. It can be seen that LCA can help to find homogeneous subgroups and the association between the groups and some certain distal variables.

Our study also fits a three-category latent class variable according to the heterogeneity of HRBs. The result is consistent with the results of some other researches. Zlotnick [54] believed that some characteristics of complex posttraumatic stress disorder(PTSD) should coexist with internalized indicators (for example, ineffectiveness, shame, depression/ despair, social withdrawal, and physical discomfort), while other characteristics should be more related to externalization indicators (for example, self-destructive behavior, impulsive actions, and hostility). Miller’s research [55] indicated internalization and externalization show higher scores on complex PTSD rather than simple PTSD in a series of scales measuring the core concept of PTSD. These findings highlight the heterogeneous of patients with complex adverse experiences. The practice of the LCA method in these researches suggests that the latent category method can better take into account the consistency within group and the difference between groups. The attempts of these studies using the LCA method suggest that the LCA method has great potential in the field of exploring adolescent behavior co-occurrence and heterogeneity.

Studies have suggested that the effects of CSA are cumulative. Exposure to high levels or multiple forms of CSA experiences may have more harmful and more reversible effects [23]. In our study, the results of association analysis showed that, adolescents presented diverse patterns of HRBs according to different patterns of CSA experience. Compared to the students classified as “Low HRBs”, those classified as “externalizing HRBs” were associated with a high probability of “Verbal or exhibitionism CSA”, and a similar association was evident among those in “internalizing HRBs”. Meanwhile, compared with “Low CSAs”, those in “high multiple CSAs”, have the strongest association with both “externalizing HRBs” and “internalizing HRBs”. All these associated directions were positive and statistically significant, which are consistent with the research hypothesis. Existing studies suggest that children with CSA experiences show higher risk of engaging in health-risk behaviors, such as drinking, smoking, gambling, and suicidal ideation [38, 49, 56, 57], or being less ambitious, having fewer friends, and having lower self-esteem [10, 58], and then their lives may have been traumatically disrupted by these behaviors [12,13,14,15,16, 59]. Compared with those who have received low or no abuse, multiple victimization are associated with an increased risk of externalization problems [60] or internalizing problems [61,62,63,64,65]. Our results not only confirmed the relation between CSA and risk behaviors, but also showed a dose-response association between the amount of ACE and behavioral and health issues, which were consistent with some of the few current studies, which means multiple victims are more likely to have behavioral or psychological problems. Sexual assault or forced penetrated abuse is more likely to associate with more negative long-term consequences [66,67,68] compared with non-forced or non-penetrated sexual abuse. More severe forms of sexual abuse are more destructive because they reinforce feelings of helplessness, powerlessness and self-blame [69]. Our study supports previous studies that multiple CSA experiences may lead to a decrease in the overall perception of one’s health [70] and an increase in negative health behaviors [71]. Abuse prevalence are believed to undermine the ability of adolescents to complete key developmental tasks, promoting them to use ineffective or even harmful coping strategies. The difference in population patterns can help researchers understand the homogeneity and heterogeneity between individuals, so as to provide a basis for the design and implementation of targeted intervention measures for different subtypes of victimization.

Previous studies showed that boys and girls who experienced CSA may have similar physical and mental health sequelae [24, 72, 73]. Our study showed that some of the association strength differed between genders. The associations were nearly twice as strong in girls as it was in boys (male OR = 3.08, 95%CI 1.48–6.40 vs female OR = 6.05, 95%CI 3.73–9.80). The results suggest that boys and girls may differ in how they respond to negative events [74, 75]. Females tend to internalize stress into shyness, shame, guilt, sadness, and self-hostility [76], while males are more likely to respond with externalizing behaviors such as problematic alcohol use or committing violent acts [77]. However, there are controversies in existing studies on the impact of gender on the associations of CSA and HRB. Coohey [78] suggested that sexually abused boys were more likely than girls to have an internalizing behavior problem using a clinical sample, while Maikovich-Fong [61, 72] claimed that sex did not moderate the relation between abuse characteristics and youth emotional and behavioral problems. These findings highlight the importance of gender in the field of CSA research.

Limitations

This study has common limitations of self-reported retrospective research. The possibility of underreport or underestimation cannot be ruled out. Due to the strict sampling frame and the high response rate of this study, volunteer bias should be minimal. Different studies have used different CSA items to define CSA, which largely affects the estimated prevalence [42]. The items of CSA experience measured in this study was embedded in a large national study. Due to the limitation of the investigation duration, the simplified scale adapted based on existing CSA measurements was not a generally accepted scale. It is difficult to avoid the impact of measurement bias and recall bias using self-reporting to collect data, which limit future comparisons between our results and other studies using different measurements. The cross-sectional design of our study was not capable of capturing chronological or cause-sequence order. The respondents were between 10 and 18 years old and had not yet experienced the full adolescence, so many of the abuse experiences or health risk behaviors will not yet have crystallized. Our findings about school students may not be extended to out-of-school children or CSA from peers. For these subpopulations, CSA prevalence or patterns may be different. Therefore, these are also possible causes of CSA underestimation. Application of LCA enables us to identify and categorize the target population into heterogeneity subgroups. The misclassification assigned to each latent class is reasonable and the probabilities of classification error were fairly accepted. However, uncertainty [79] in model class membership still exists. The results should be cautiously interpreted. Rare prevalence of CSA experiences, underestimation and potential misclassification may cause smaller sample cells and may reduce the statistical power. Due to the covert and sensitivity nature of the survey contents, randomization, restrictions or matching were not applied to controlled potential confounding, while we adopted gender stratification and multivariate dual LCA regression model to controll measured confounders, in order to reduce potential bias. Our analysis was based on a limited number of covariates, thus revealing a broader range of confounding factors that may also influence the association between CSA and HRB is in need. Further considerations should be tested in cooperating broader potential mediators, moderators and their possible impacts on adolescent adverse experiences or behaviors.

Conclusion and implication

Notwithstanding the limitations, our study contributes to the scarce literature on exploring prevalence and co-occurrence subgroups of Chinese school teens reporting sexual abuse and examines the association between CSA and HRB patterns. The prevalence of experiencing at least one form of CSAs is 12.9%. Non-contact CSAs are the most common forms. The co-occurrence and heterogeneity of the experiences play a role in the association of CSA and HRB. Students who have experienced multiple-victimization of CSA have a higher probability of HRB risk, especially “depression, despair, or extreme anxiety”, “suicidal ideation” and “suicidal behavior”. The findings emphasize the importance of considering heterogeneity and co-occurrence in studies of adverse childhood experiences. The analysis could improve understandings of the potential vulnerabilities. Although contact or non-contact CSA are considered public health problems, multi-victimization has more significantly negative effects. Children with multiple abuse experiences should be paid more attention to for risk behavior interventions. The results could provide new evidence and research directions for association between CSA and HRB, and help revise the interventions for different subgroups of victims.

Suggestions for future studies

This study aims to explore the potential mechanism and to provide new clues and research directions for the association between CSA and HRB. Future research that replicates this research is needed to generalize these findings to similar or different samples of adolescents, and to examine differences between subgroups with specific risk factors that have the potential to mediate or moderate the association, which could help develop more nuanced interventions addressing adverse experiences and risk behaviors among at-risk adolescents.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Abbreviations

CSA:

Child sexual abuse

HRB:

Health risk behaviors

OR:

Odds Ratio

CI:

Confidence Interval

LCA:

Latent class analysis

ISPCAN:

International Society for Prevention of Child Abuse and Neglect

AIC:

Akaike Information Criterion

BIC:

Bayesian Information Criterion

aBIC:

Adjusted Bayesian Information Criterion

References

  1. 1.

    World Health Organization. Report of the consultation on child abuse prevention. Geneva: World Health Organization; 1999. https://apps.who.int/iris/handle/10665/65900.

  2. 2.

    Hughes K, Bellis MA, Hardcastle KA, Sethi D, Butchart A, Mikton C, et al. The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis. Lancet Public Health. 2017;2(8):e356–e66. https://doi.org/10.1016/S2468-2667(17)30118-4.

    Article  PubMed  Google Scholar 

  3. 3.

    Organization WH. Preventing child maltreatment: a guide to taking action and generating evidence. 2006.

    Google Scholar 

  4. 4.

    Stoltenborgh M, van Ijzendoorn MH, Euser EM, Bakermans-Kranenburg MJ. A global perspective on child sexual abuse: meta-analysis of prevalence around the world. Child Maltreat. 2011;16(2):79–101. https://doi.org/10.1177/1077559511403920.

    Article  PubMed  Google Scholar 

  5. 5.

    Barth J, Bermetz L, Heim E, Trelle S, Tonia T. The current prevalence of child sexual abuse worldwide: a systematic review and meta-analysis. Int J Public Health. 2013;58(3):469–83. https://doi.org/10.1007/s00038-012-0426-1.

    CAS  Article  PubMed  Google Scholar 

  6. 6.

    Pereda N, Guilera G, Forns M, Gomez-Benito J. The prevalence of child sexual abuse in community and student samples: a meta-analysis. Clin Psychol Rev. 2009;29(4):328–38. https://doi.org/10.1016/j.cpr.2009.02.007.

    Article  PubMed  Google Scholar 

  7. 7.

    Green JG, McLaughlin KA, Berglund PA, Gruber MJ, Sampson NA, Zaslavsky AM, et al. Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication I: associations with first onset of DSM-IV disorders. Arch Gen Psychiatry. 2010;67(2):113–23. https://doi.org/10.1001/archgenpsychiatry.2009.186.

    Article  PubMed  PubMed Central  Google Scholar 

  8. 8.

    Hamby S, Grych J. The web of violence: Exploring connections among different forms of interpersonal violence and abuse; 2013. https://doi.org/10.1007/978-94-007-5596-3.

    Book  Google Scholar 

  9. 9.

    Widom CS, DuMont K, Czaja SJ. A prospective investigation of major depressive disorder and comorbidity in abused and neglected children grown up. Arch Gen Psychiatry. 2007;64(1):49–56. https://doi.org/10.1001/archpsyc.64.1.49.

    Article  PubMed  Google Scholar 

  10. 10.

    Al-Fayez GA, Ohaeri JU, Gado OM. Prevalence of physical, psychological, and sexual abuse among a nationwide sample of Arab high school students: association with family characteristics, anxiety, depression, self-esteem, and quality of life. Soc Psychiatry Psychiatr Epidemiol. 2012;47(1):53–66. https://doi.org/10.1007/s00127-010-0311-2.

    Article  PubMed  Google Scholar 

  11. 11.

    Steine IM, Krystal JH, Nordhus IH, Bjorvatn B, Harvey AG, Eid J, et al. Insomnia, nightmare frequency, and nightmare distress in victims of sexual abuse. J Interpers Viol. 2011;27(9):1827–43. https://doi.org/10.1177/0886260511430385.

    Article  Google Scholar 

  12. 12.

    Boduszek D, Hyland P, Bourke A. An investigation of the role of personality, familial, and peer-related characteristics in homicidal offending using retrospective data. J Crim Psychol. 2012;2(2):96–106. https://doi.org/10.1108/20093821211264414.

    Article  Google Scholar 

  13. 13.

    Dhingra K, Boduszek D, Sharratt K. Victimization profiles, non-suicidal self-injury, suicide attempt, and post-traumatic stress disorder symptomology: application of latent class analysis. J Interpers Viol. 2016;31(14):2412–29. https://doi.org/10.1177/0886260515576967.

    Article  Google Scholar 

  14. 14.

    Lundberg-Love PK, Marmion S, Ford K, Geffner R, Peacock L. The long-term consequences of childhood incestuous victimization upon adult Women's psychological symptomatology. J Child Sex Abuse. 2008;1(1):81–102. https://doi.org/10.1300/J070v01n01_06.

    Article  Google Scholar 

  15. 15.

    Mills R, Kisely S, Alati R, Strathearn L, Najman JM. Child maltreatment and cannabis use in young adulthood: a birth cohort study. Addiction. 2017;112(3):494–501. https://doi.org/10.1111/add.13634.

    Article  PubMed  Google Scholar 

  16. 16.

    Vachon DD, Krueger RF, Rogosch FA, Cicchetti D. Assessment of the harmful psychiatric and behavioral effects of different forms of child maltreatment. JAMA Psychiatry. 2015;72(11):1135–42. https://doi.org/10.1001/jamapsychiatry.2015.1792.

    Article  PubMed  PubMed Central  Google Scholar 

  17. 17.

    Chan KL. Victimization and poly-victimization among school-aged Chinese adolescents: prevalence and associations with health. Prev Med. 2013;56(3–4):207–10. https://doi.org/10.1016/j.ypmed.2012.12.018.

    Article  PubMed  Google Scholar 

  18. 18.

    Hesketh T, Zhen Y, Lu L, Dong ZX, Jun YX, Xing ZW. Stress and psychosomatic symptoms in Chinese school children: cross-sectional survey. Arch Dis Child. 2010;95(2):136–40. https://doi.org/10.1136/adc.2009.171660.

    Article  PubMed  Google Scholar 

  19. 19.

    Ozer EJ, McDonald KL. Exposure to violence and mental health among Chinese American urban adolescents. J Adolesc Health. 2006;39(1):73–9. https://doi.org/10.1016/j.jadohealth.2005.09.015.

    Article  PubMed  Google Scholar 

  20. 20.

    McLean CP, Morris SH, Conklin P, Jayawickreme N, Foa EB. Trauma characteristics and posttraumatic stress disorder among adolescent survivors of childhood sexual abuse. J Fam Violence. 2014;29(5):559–66. https://doi.org/10.1007/s10896-014-9613-6.

    Article  PubMed  PubMed Central  Google Scholar 

  21. 21.

    Moore EE, Romaniuk H, Olsson CA, Jayasinghe Y, Carlin JB, Patton GC. The prevalence of childhood sexual abuse and adolescent unwanted sexual contact among boys and girls living in Victoria, Australia. Child Abuse Negl. 2010;34(5):379–85. https://doi.org/10.1016/j.chiabu.2010.01.004.

    Article  PubMed  Google Scholar 

  22. 22.

    Higgins DJ, McCabe MP. Multiple forms of child abuse and neglect: adult retrospective reports. Aggress Violent Behav. 2001;6(6):547–78. https://doi.org/10.1016/s1359-1789(00)00030-6.

    Article  Google Scholar 

  23. 23.

    Finkelhor D, Ormrod RK, Turner HA. Poly-victimization: a neglected component in child victimization. Child Abuse Negl. 2007;31(1):7–26. https://doi.org/10.1016/j.chiabu.2006.06.008.

    Article  PubMed  Google Scholar 

  24. 24.

    Paolucci EO, Genuis ML, Violato C. A meta-analysis of the published research on the effects of child sexual abuse. J Psychol. 2001;135(1):17–36. https://doi.org/10.1080/00223980109603677.

    CAS  Article  PubMed  Google Scholar 

  25. 25.

    Trickett PK. Noll, J. G., and Putnam, F. W. the impact of sexual abuse on female development: lessons from a multigenerational, longitudinal research study. Dev Psychopathol. 2011;23(2):453–76. https://doi.org/10.1017/S0954579411000174.

    Article  PubMed  PubMed Central  Google Scholar 

  26. 26.

    Ford JD, Elhai JD, Connor DF, Frueh BC. Poly-victimization and risk of posttraumatic, depressive, and substance use disorders and involvement in delinquency in a national sample of adolescents. J Adolesc Health. 2010;46(6):545–52. https://doi.org/10.1016/j.jadohealth.2009.11.212.

    Article  PubMed  Google Scholar 

  27. 27.

    Ford JD, Grasso DJ, Hawke J, Chapman JF. Poly-victimization among juvenile justice-involved youths. Child Abuse Negl. 2013;37(10):788–800. https://doi.org/10.1016/j.chiabu.2013.01.005.

    Article  PubMed  Google Scholar 

  28. 28.

    Bakk Z, Tekle FB, Vermunt JK. Estimating the association between latent class membership and external variables using Bias-adjusted three-step approaches. Sociol Methodol. 2013;43(1):272–311. https://doi.org/10.1177/0081175012470644.

    Article  Google Scholar 

  29. 29.

    Finkelhor D, Ormrod R, Turner H, Hamby SL. The victimization of children and youth: a comprehensive, national survey. Child Maltreat. 2005;10(1):5–25. https://doi.org/10.1177/1077559504271287.

    Article  PubMed  Google Scholar 

  30. 30.

    Davis JL, Petretic-Jackson PA. The impact of child sexual abuse on adult interpersonal functioning: A review and synthesis of the empirical literature. Aggression Viol Behav. 2000;5:291–328.

    Article  Google Scholar 

  31. 31.

    Jackson N, Denny S, Sheridan J, Fleming T, Clark T, Teevale T, et al. Predictors of drinking patterns in adolescence: a latent class analysis. Drug Alcohol Depend. 2014;135:133–9. https://doi.org/10.1016/j.drugalcdep.2013.11.021.

    Article  PubMed  Google Scholar 

  32. 32.

    Reboussin BA, Song EY, Shrestha A, Lohman KK, Wolfson M. A latent class analysis of underage problem drinking: evidence from a community sample of 16-20 year olds. Drug Alcohol Depend. 2006;83(3):199–209. https://doi.org/10.1016/j.drugalcdep.2005.11.013.

    Article  PubMed  Google Scholar 

  33. 33.

    Lau JTF, Liu JLY, Yu A, Wong CK. Conceptualization, reporting and underreporting of child abuse in Hong Kong. Child Abuse Neglect. 1999;23(11):1159–74. https://doi.org/10.1016/S0145-2134(99)00075-7 Elsevier Ltd.

    CAS  Article  PubMed  Google Scholar 

  34. 34.

    Sun YP, Zhang B, Dong ZJ, Yi MJ, Sun DF, Shi SS. Psychiatric state of college students with a history of childhood sexual abuse. World J Pediatr. 2008;4(4):289–94. https://doi.org/10.1007/s12519-008-0052-4.

    Article  PubMed  Google Scholar 

  35. 35.

    Wong ML, Chan RK, Koh D, Tan HH, Lim FS, Emmanuel S, et al. Premarital sexual intercourse among adolescents in an Asian country: multilevel ecological factors. Pediatrics. 2009;124(1):e44–52. https://doi.org/10.1542/peds.2008-2954.

    Article  PubMed  Google Scholar 

  36. 36.

    Chan KL. Correlates of childhood sexual abuse and intimate partner sexual victimization. Partn Abus. 2011;2(3):365–81. https://doi.org/10.1891/1946-6560.2.3.365.

    Article  Google Scholar 

  37. 37.

    Hulme PA. Retrospective measurement of childhood sexual abuse: a review of instruments. Child Maltreat. 2004;9(2):201–17. https://doi.org/10.1177/1077559504264264.

    Article  PubMed  Google Scholar 

  38. 38.

    Chen J, Dunne MP, Han P. Child sexual abuse in China: a study of adolescents in four provinces. Child Abuse Negl. 2004;28(11):1171–86. https://doi.org/10.1016/j.chiabu.2004.07.003.

    Article  PubMed  Google Scholar 

  39. 39.

    Ji K, Finkelhor D, Dunne M. Child sexual abuse in China: a meta-analysis of 27 studies. Child Abuse Negl. 2013;37(9):613–22. https://doi.org/10.1016/j.chiabu.2013.03.008.

    Article  PubMed  Google Scholar 

  40. 40.

    Muthén LK, Muthén BO. Mplus User’s Guide. 8th ed. Los Angeles: Muthén & Muthén; 1998-2017.

    Google Scholar 

  41. 41.

    Landolt MA, Schnyder U, Maier T, Mohler-Kuo M. The harm of contact and non-contact sexual abuse: health-related quality of life and mental health in a population sample of Swiss adolescents. Psychother Psychosom. 2016;85(5):320–2. https://doi.org/10.1159/000446810.

    Article  PubMed  Google Scholar 

  42. 42.

    Fergusson DM. & Mullen, P. E. childhood sexual abuse: an evidence based perspective. Child Abuse Rev. 1999;10(5):367–8. https://doi.org/10.1002/car.674.

    Article  Google Scholar 

  43. 43.

    Luo Y, Parish WL, Laumann EO. A population-based study of childhood sexual contact in China: prevalence and long-term consequences. Child Abuse Negl. 2008;32(7):721–31. https://doi.org/10.1016/j.chiabu.2007.10.005.

    Article  PubMed  PubMed Central  Google Scholar 

  44. 44.

    Huang G, Zhang Y, Momartin S, Huang X, Zhao L. Child sexual abuse, coping strategies and lifetime posttraumatic stress disorder among female inmates. Int J Prison Health. 2008;4(1):54–63. https://doi.org/10.1080/17449200701875840.

    Article  PubMed  Google Scholar 

  45. 45.

    Chen JQ, Dunne MP, Wang XW. Prevalence of childhood sexual abuse among female students in a senior high school. Chin J School Health. 2002;23:108–10.

    Google Scholar 

  46. 46.

    Chen JQ, Dunne MP, Wang XW. Childhood sexual abuse: an investigation among 239 male high school students. Chin Ment Health J. 2003;17(5):345–7.

    Google Scholar 

  47. 47.

    Li CQ. Analyzing the family factor on the childhood sexual abuse among college student. Chin J School Health. 2008;29:536–7.

    Google Scholar 

  48. 48.

    Sun YP, Dong ZJ, Yi MJ, Sun DF. Childhood sexual abuses among 1307 adult students and analysis on results of symptom Checklist-90 test. Zhonghua er ke za zhi Chin J Pediatr. 2006;44(1):21–5.

    Google Scholar 

  49. 49.

    Lin D, Li X, Fan X, Fang X. Child sexual abuse and its relationship with health risk behaviors among rural children and adolescents in Hunan, China. Child Abuse Neglect. 2011;35(9):680–7. https://doi.org/10.1016/j.chiabu.2011.05.006.

    Article  PubMed  Google Scholar 

  50. 50.

    Chen JQ, Ma Y, Cheng Y, Chen B, Liang Y. A survey on the occurrence and influencing factors of childhood sexual abuse among adolescent students. Chin J Hum Sex. 2006;15(8):3–7.

    Google Scholar 

  51. 51.

    Munro K. Sexual abuse: definitions, perpetrators, victims and effects; 2000.

    Google Scholar 

  52. 52.

    Shevlin M, Murphy S, Elklit A, Murphy J, Hyland P. Typologies of child sexual abuse: an analysis of multiple abuse acts among a large sample of Danish treatment-seeking survivors of childhood sexual abuse. Psychol Trauma. 2018;10(3):263–9. https://doi.org/10.1037/tra0000268.

    Article  PubMed  Google Scholar 

  53. 53.

    French BH, Bi Y, Latimore TG, Klemp HR, Butler EE. Sexual victimization using latent class analysis: exploring patterns and psycho-behavioral correlates. J Interpers Viol. 2014;29(6):1111–31. https://doi.org/10.1177/0886260513506052.

    Article  Google Scholar 

  54. 54.

    Zlotnick C, Warshaw M, Shea MT, Allsworth J, Pearlstein T, Keller MB. Chronicity in posttraumatic stress disorder (PTSD) and predictors of course of comorbid PTSD in patients with anxiety disorders. J Trauma Stress. 1999;12(1):89–100. https://doi.org/10.1023/A:1024746316245.

    CAS  Article  PubMed  Google Scholar 

  55. 55.

    Miller MW, Fogler JM, Wolf EJ, Kaloupek DG, Keane TM. The internalizing and externalizing structure of psychiatric comorbidity in combat veterans. J Trauma Stress. 2008;21(1):58–65. https://doi.org/10.1002/jts.20303.

    Article  PubMed  PubMed Central  Google Scholar 

  56. 56.

    Chen JQ, Han P, Dunne MP. Child sexual abuse: a study among 892 female students of a medical school. Zhonghua Er Ke Za Zhi. 2004;42(1):39–43.

    PubMed  Google Scholar 

  57. 57.

    Zhu Q, Gao E, Cheng Y, Chuang YL, Zabin LS, Emerson MR, et al. Child sexual abuse and its relationship with health risk behaviors among adolescents and young adults in Taipei. Asia Pac J Public Health. 2015;27(6):643–51. https://doi.org/10.1177/1010539515573075.

    Article  PubMed  Google Scholar 

  58. 58.

    Swanston HY, Plunkett AM, O’Toole BI, Shrimpton S, Parkinson PN, Oates RK. Nine years after child sexual abuse. Child Abuse Negl. 2003;27(8):967–84. https://doi.org/10.1016/s0145-2134(03)00143-1.

    Article  PubMed  Google Scholar 

  59. 59.

    Dube SR, Anda RF, Whitfield CL, Brown DW, Felitti VJ, Dong M, et al. Long-term consequences of childhood sexual abuse by gender of victim. Am J Prev Med. 2005;28(5):430–8. https://doi.org/10.1016/j.amepre.2005.01.015.

    Article  PubMed  Google Scholar 

  60. 60.

    Lyons J, Romano E. Childhood sexual abuse profiles and psychological functioning in adult males. J Child Sex Abuse. 2019;28(5):544–63. https://doi.org/10.1080/10538712.2018.1563263.

    Article  Google Scholar 

  61. 61.

    Rapsey CM, Scott KM, Patterson T. Childhood sexual abuse, poly-victimization and internalizing disorders across adulthood and older age: findings from a 25-year longitudinal study. J Affect Disord. 2019;244:171–9. https://doi.org/10.1016/j.jad.2018.10.095.

    Article  PubMed  Google Scholar 

  62. 62.

    Chou KL. Childhood sexual abuse and psychiatric disorders in middle-aged and older adults: evidence from the 2007 adult psychiatric morbidity survey. J Clin Psychiatry. 2012;73(11):e1365–71. https://doi.org/10.4088/JCP.12m07946.

    Article  PubMed  Google Scholar 

  63. 63.

    Draper B, Pfaff JJ, Pirkis J, Snowdon J, Lautenschlager NT, Wilson I, et al. Long-term effects of childhood abuse on the quality of life and health of older people: results from the depression and early prevention of suicide in general practice project. J Am Geriatr Soc. 2008;56(2):262–71. https://doi.org/10.1111/j.1532-5415.2007.01537.x.

    Article  PubMed  Google Scholar 

  64. 64.

    Ege MA, Messias E, Thapa PB, Krain LP. Adverse childhood experiences and geriatric depression: results from the 2010 BRFSS. Am J Geriatr Psychiatry. 2015;23(1):110–4. https://doi.org/10.1016/j.jagp.2014.08.014.

    Article  PubMed  Google Scholar 

  65. 65.

    Raposo SM, Mackenzie CS, Henriksen CA, Afifi TO. Time does not heal all wounds: older adults who experienced childhood adversities have higher odds of mood, anxiety, and personality disorders. Am J Geriatr Psychiatry. 2014;22(11):1241–50. https://doi.org/10.1016/j.jagp.2013.04.009.

    Article  PubMed  Google Scholar 

  66. 66.

    Steine IM, Winje D, Krystal JH, Bjorvatn B, Milde AM, Gronli J, et al. Cumulative childhood maltreatment and its dose-response relation with adult symptomatology: findings in a sample of adult survivors of sexual abuse. Child Abuse Negl. 2017;65:99–111. https://doi.org/10.1016/j.chiabu.2017.01.008.

    Article  PubMed  Google Scholar 

  67. 67.

    Briere J, Elliott DM. Prevalence and psychological sequelae of self-reported childhood physical and sexual abuse in a general population sample of men and women. Child Abuse Negl. 2003;27(10):1205–22. https://doi.org/10.1016/j.chiabu.2003.09.008.

    Article  PubMed  Google Scholar 

  68. 68.

    Cutajar MC, Mullen PE, Ogloff JRP, Thomas SD, Wells DL, Spataro J. Suicide and fatal drug overdose in child sexual abuse victims: a historical cohort study. Med J Aust. 2010;192(4):184–7. https://doi.org/10.5694/j.1326-5377.2010.tb03475.x.

    Article  PubMed  Google Scholar 

  69. 69.

    Hahm HC, Lee Y, Ozonoff A, Van Wert MJ. The impact of multiple types of child maltreatment on subsequent risk behaviors among women during the transition from adolescence to young adulthood. J Youth Adolesc. 2010;39(5):528–40. https://doi.org/10.1007/s10964-009-9490-0.

    Article  PubMed  Google Scholar 

  70. 70.

    Bassani DG, Palazzo LS, Beria JU, Gigante LP, Figueiredo ACL, Aerts DRGC, et al. Child sexual abuse in southern Brazil and associated factors: a population-based study. BMC Public Health. 2009;9(1):133. https://doi.org/10.1186/1471-2458-9-133.

    Article  PubMed  PubMed Central  Google Scholar 

  71. 71.

    Brown D. Exposure to physical and sexual violence and adverse health behaviours in African children: results from the global school-based student health survey. Bull World Health Organ. 2009;87(6):447–55. https://doi.org/10.2471/blt.07.047423.

    Article  PubMed  PubMed Central  Google Scholar 

  72. 72.

    Maikovich-Fong AK, Jaffee SR. Sex differences in childhood sexual abuse characteristics and victims' emotional and behavioral problems: findings from a national sample of youth. Child Abuse Negl. 2010;34(6):429–37. https://doi.org/10.1016/j.chiabu.2009.10.006.

    Article  PubMed  PubMed Central  Google Scholar 

  73. 73.

    Rhodes AE, Boyle MH, Tonmyr L, Wekerle C, Goodman D, Leslie B, et al. Sex differences in childhood sexual abuse and suicide-related behaviors. Suicide Life Threat Behav. 2011;41(3):235–54. https://doi.org/10.1111/j.1943-278X.2011.00025.x.

    Article  PubMed  Google Scholar 

  74. 74.

    Hurrelmann KR, M. Risk behaviour in adolescence: the relationship between developmental and health problems. J Public Health. 2006;14(1):20–8. https://doi.org/10.1007/s10389-005-0005-5.

    Article  Google Scholar 

  75. 75.

    Morimoto Y, Sharma A. Long-term outcomes of verbal aggression. J Emot Abus. 2004;4(2):71–99. https://doi.org/10.1300/J135v04n02_04.

    Article  Google Scholar 

  76. 76.

    Broidy L, Agnew R. Gender and crime: a general strain theory perspective. J Res Crime Delinq. 2016;34(3):275–306. https://doi.org/10.1177/0022427897034003001.

    Article  Google Scholar 

  77. 77.

    Bender K. Why do some maltreated youth become juvenile offenders? Child Youth Serv Rev. 2010;32(3):466–73. https://doi.org/10.1016/j.childyouth.2009.10.022.

    Article  Google Scholar 

  78. 78.

    Coohey C. Gender differences in internalizing problems among sexually abused early adolescents. Child Abuse Negl. 2010;34(11):856–62. https://doi.org/10.1016/j.chiabu.2010.05.001.

    Article  PubMed  Google Scholar 

  79. 79.

    Bray BC, Lanza ST, Tan X. Eliminating Bias in classify-analyze approaches for latent class analysis. Struct Equ Modeling. 2015;22(1):1–11. https://doi.org/10.1080/10705511.2014.935265.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

The authors thank site coordinators of the seven provinces for their hard work. The authors also appreciate the cooperation of the students involved.

Funding

The research was funded by the Ministry of Science and Technology, P.R. China (Award no. 2012BAI32B02) and the Clinical Research Project of Shanghai Municipal Commission of Health and Family Planning(Award no: 201840095). The funding bodies have played no role in the design of the study, collection of data, analysis of data, and interpreting the results or in writing the manuscript or in the decision to submit the manuscript for publication. The content of this publication is only the responsibility of the authors.

Author information

Affiliations

Authors

Contributions

WZ, CL, XT, SZ, and SL conceptualized and designed the study; XZ, YM, QL, SL, and SZ collected and cleaned the data; YZ carried out statistical analysis and drafted the manuscript; YZ and QL interpreted the results; CL, WZ, and XT contributed to discussion; and all authors read and approved the final manuscript.

Corresponding author

Correspondence to Weijin Zhou.

Ethics declarations

Ethics approval and consent to participate

This study was reviewed and approved by the institutional review board of Shanghai Institute of Planned Parenthood Research(reference number: 2012–01). All survey participants and their parents read and agreed to a written informed consent document prior to completing the survey and are given the right to skip any part of the survey or question. All methods were carried out in accordance with the ethical principles of the Declaration of Helsinki 1964.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher’s Note

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

Supplementary Information

Additional file 1:

The table shows CSA prevalence stratified by the participant demographic characteristics.

Rights and permissions

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 http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Zhang, Y., Zuo, X., Mao, Y. et al. Co-occurrence subgroups of child sexual abuse, health risk behaviors and their associations among secondary school students in China. BMC Public Health 21, 1139 (2021). https://doi.org/10.1186/s12889-021-11199-5

Download citation

Keywords

  • Middle school students
  • Child sexual abuse
  • Health risk behaviors
  • Latent class analysis