Adverse childhood experiences (ACEs) are a series of potentially traumatic experiences that occur before the age of 18 (Sosnowski et al., 2021), including harms that directly affect children (e.g., abuse and neglect, or extrafamilial violence) and indirectly affect children’s life circumstances (e.g., family dysfunction, or parental absence) (Hughes et al., 2017). The cumulative risk perspective emphasizes the importance of the number of ACEs, stressing that the greater the number of adversities experienced by an individual, the greater the negative impacts (Evans et al., 2013). Deviant peer affiliation is conceptualized as selective participation in deviant friendships while peer interaction and encouragement of criminal behaviors occur (Bao et al., 2022). Numerous empirical studies have shown that adverse childhood experiences are associated with negative life outcomes, such as a high risk of developmental, social, or behavioral delays (Cprek et al., 2020), substance abuse behaviors (Forster et al., 2018), and externalizing problem behaviors (Hunt et al., 2017). Similarly, the overwhelming correlation between deviant peer affiliation and externalizing problem behaviors has been confirmed (Zimmerman and Messner, 2011). Despite this, there has been relatively little research on how adverse childhood experiences affect delinquent peer affiliation (Jones et al., 2023). One of the few studies conducted by Trinidad demonstrated that individuals who experience family adversity during childhood were more likely to engage in deviant peer affiliation during adolescence and early adulthood (Trinidad, 2021). Jones et al. (2023) suggested that as children’s exposure to ACEs increases (by age 5), the likelihood of reporting association with delinquent peers increases.

In summary, there is some evidence to suggest a possible pattern between adverse childhood experiences and deviant peer affiliation. Most of the current research is conducted in general population groups, and it is worth considering whether their patterns of peer interactions are fully consistent with those of at-risk groups.

Relative deprivation is a judgment that an individual is worse off compared to certain standards, and is accompanied by feelings of anger and resentment (Smith et al., 2012; Li et al., 2023). According to the theoretical model of relative deprivation described by Smith and colleagues, relative deprivation involves three steps: cognitive comparisons, cognitive appraisals, and justice-related affect (Smith et al., 2012). During adolescence, individuals’ peer interactions increase, providing a basis for social comparisons (McGloin and Thomas, 2019). Adolescents often believe that they deserve positive childhood experiences and that they are not responsible for the occurrence of adverse childhood experiences. Consequently, those who have experienced adversity may perceive themselves as disadvantaged after cognitive appraisal, and develop a sense of relative deprivation, which in turn leads to behavioral consequences resulting from that sense of relative deprivation (Li et al., 2023). Therefore, it is reasonable to expect that feelings of relative deprivation play a mediating role in ACEs and deviant peer affiliation interactions. Adolescents who experience more ACEs may develop higher levels of relative deprivation, which leads to more deviant peer affiliation.

There are two reasons to support the argument for relative deprivation playing a mediating role. Firstly, adolescents who experience ACEs may experience higher levels of relative deprivation. While there is no direct evidence linking ACEs to feelings of relative deprivation, there has been a significant body of research demonstrating that individuals who experience ACEs feel higher levels of shame, more envy, and are more susceptible to psychiatric disorders such as depression and anxiety than controls, as well as having stronger motivation to make social comparisons (Dehon and Weems, 2010; Xiang et al., 2018, 2020). Meanwhile, comparisons with others are a key mechanism in the creation of relative deprivation (Smith et al., 2012). Furthermore, some studies have demonstrated that individuals who lacked social support or experienced childhood maltreatment reported higher levels of relative deprivation (Zhang and Tao, 2013; Li et al., 2023). Second, relative deprivation may be positively correlated with deviant peer affiliation. Relative deprivation may lead to negative outcomes such as anger or aberrant behavior in individuals. Research based on the Chinese context suggested that relative deprivation has a positive effect on violent crime (Smith et al., 2012; Song et al., 2022). Through three experimental studies, researchers have shown that participants in a relative deprivation condition exhibit higher levels of aggression than individuals who get along in a relative contentment condition (Greitemeyer and Sagioglou, 2017). Similarly, it has been demonstrated that a sense of relative deprivation is significantly and positively associated with deviant peer affiliation (Li et al., 2023).

Although ACE is an important predictor of relative deprivation, it is crucial to consider the impact of adolescents’ cognitive development at different stages, as this is a process that develops over time (Callan et al., 2015). The theory of relative deprivation posits that an individual’s sense of relative deprivation involves three integral parts: comparing oneself to others, recognizing that one is disadvantaged, and experiencing negative emotions (Smith et al., 2012). Perspectives from developmental psychology suggest that adolescence is a period of active social comparison (Harter, 2015). Firstly, upon entering secondary school, students are exposed to more information regarding social comparison due to a more diverse peer group and a developmental stage that places greater emphasis on peers rather than family (Yoon et al., 2019; Huang and Zhu, 2020). Secondly, many studies have pointed out that the transition to secondary school is marked by a shift in educational emphasis, which is reflected in teachers’ greater focus on social comparisons of students, such as publishing grades (Fredricks and Eccles, 2002; Gürel et al., 2020). At the same time, research has shown that self-evaluation fluctuates during early adolescence (Siffert et al., 2012). Meta-studies have indicated that the stability of self-evaluation increases throughout adolescence and adulthood (Trzesniewski et al., 2003), with one possible explanation being changes in significant others during early adolescence (Siffert et al., 2012). In addition, evidence from experimental studies suggested that there is an increase in adolescents’ ability to regulate their emotions from early to mid-adolescence (Theurel and Gentaz, 2018). Therefore, it is reasonable to hypothesize that age is a moderating variable for the relative deprivation of ACEs. According to the researcher’s recommendations, adolescents can be categorized into early adolescence (age 10–14 years) and late adolescence (age 15–19 years) based on their age (Irwin et al., 2002). In early adolescence, ACEs were stronger predictors of relative deprivation; in late adolescence, ACEs predicted relative poverty weakly. Different ages of adolescents may influence not only the relationship between ACEs and relative deprivation, but also the indirect role of ACEs in influencing deviant peer affiliation through relative deprivation. Specifically, the indirect effect of ACEs influencing deviant peer affiliation through relative deprivation was stronger in early adolescence; it was weaker in late adolescence.

Thus, this study aims to examine the mediating roles of relative deprivation between ACEs and deviant peer affiliation, as well as the moderating role of age in this process, using a sample of delinquent adolescents in China (see Figure 1).

The study sample consisted of 542 adolescents from five Specialized Schools in China. In China, as stipulated in the Law on the Prevention of Juvenile Delinquency, specialized schools are designed for three categories of persons: minors who have committed criminal acts, minors who have committed serious violations of public order, and minors who have committed general violations of public order (Leng and Jiao, 2023; Zhang and Yao, 2023). Participants ranged in age from 11 to 18, with an average of 14.50 years (SD = 1.36). Among them, 78.20% (424/542) were males, 21.8% (118/542) were females; 17.70% have no siblings. 50.20% (272/542) could be classified as early adolescence and 49.80% (270/542) as late adolescence.

The current study was approved by the Research Ethics Committee of the first author’s university. Data collection was conducted by the first author of this study. Respondents voluntarily participated in an anonymous questionnaire, and informed consent was obtained from the participants prior to commencement. Participants were informed of their right to refuse participation and to withdraw. The first author of this paper entered the classroom as a researcher after obtaining permission from the school, and the participants filled out the questionnaire in the classroom accompanied by the classroom teacher.

All data were dealt with in SPSS 27.0. We handled the missing data by mean imputation. The skewness and kurtosis of ACEs, deviant peer affiliation and relative deprivation were 0.10 and − 0.55, 0.11 and − 0.16, 0.03 and 0.08, which fell within the acceptable range (Sk < |2.0| and Ku < |7.0|; Mueller et al., 2010). The results of common method bias showed that total eigenvalues >1 were obtained without rotation. Additionally, there were 21 common factors, and the explanation rate of the first common factor was 13.41%, which was lower than 40%. Therefore, there was no serious common method deviation.

Then, the data were analyzed in three steps. First, we conducted descriptive statistics and analyzed the correlations between the variables. We also analyzed the prevalence of ACEs and statistics on deviant peer affiliation. Second, Hayes’s (2013) PROCESS 3.5 macro (Model 4) was used to examine whether relative deprivation mediated the relationship between ACEs and deviant peer affiliation. Meanwhile, the study used 5,000 bootstrap samples and corrected for bias according to 95% confidence intervals (CIs) to calculate the indirect effects of each variable. If the CI for the indirect effect did not include zero, it indicated that the indirect effect was significant at p = 0.05 (Shrout and Bolger, 2002). Third, we explored whether this mediation process was moderated by age using Hayes’s (2013) PROCESS 3.5 macro (Model 7). In all analyses, we controlled for gender as a covariates.

Firstly, 67.50% (366/542) of these 542 adolescents had the experience of being left-behind. Secondly, we tested the incidence of ACEs, the average number of ACEs experienced by adolescents was 5.07 ± 2.73 in early adolescence and 5.57 ± 2.55 in late adolescence. 97.4% (528/542) of adolescents reported one or more types of ACEs, of these, 85.8% (465/542) reported experiencing childhood abuse and neglect, 79.3% (430/542) reported experiencing extrafamilial violence, 29.5% (160/542) reported experiencing family dysfunction, and 79.0% (428/542) reported experiencing the absenteeism of a father or mother. Additionally, 99.8% (541/542) of adolescents reported one or more types of deviant peer affiliation.

Means, SDs, and correlation matrix of variables were presented in Table 1. Overall, all results were consistent with our expectations. Specifically, ACEs were positively correlated with relative deprivation and deviant peer affiliation. Furthermore, relative deprivation was positively correlated with deviant peer affiliation.

As presented in Table 2, the results showed that after controlling for gender, ACEs were positively correlated with deviant peer affiliation (b = 0.07, p < 0.001), ACEs were positively correlated with relative deprivation (b = 0.27, p < 0.001), and relative deprivation was positively correlated with deviant peer affiliation (b = 0.06, p < 0.001), thus, supporting Hypothesis 1. When relative deprivation was introduced as a mediating variable, the relationship between ACEs and deviant peer affiliation remained significant (β = 0.06, p < 0.001). Thus, relative deprivation partially mediated the relationship between ACEs and deviant peer affiliation. Furthermore, this study calculated 95% CI based on a 5,000 bootstrap resampling to further test the indirect effect, the results showed ACEs increased deviant peer affiliation through relative deprivation (b = 0.02, 95%CI [0.01, 0.03]), supporting Hypothesis 2.

As Table 3 illustrates, ACEs positively predicted relative deprivation (b = 0.57, p < 0.001). The interaction effect of ACEs and relative deprivation on age was significant (b = −0.20, p < 0.05).

To further illustrate the effect of age on the interaction between ACEs and relative deprivation (see Figure 2), the study conducted a simple slope analysis. Simple slope tests showed that the ACEs were positively associated with relative deprivation for delinquent adolescents in early adolescence (b_simple = 0.37, t = 6.01, p < 0.001). However, for those with late adolescence, the association was still significant but much weaker (b_simple = 0.17, t = 2.61, p < 0.001). Thus, Hypothesis 3(a) was supported.

In addition, this study tested conditional indirect effects as suggested by Preacher et al. (2007), which focus on examining mediating effects that are significant at different levels of moderating variables. Findings showed that relative deprivation mediated stronger effects of ACEs on deviant peer affiliation in early adolescence (b = 0.02, 95%CI [0.01, 0.03]). In late adolescence, relative deprivation mediated weaker effects in ACEs on deviant peer affiliation (b = 0.01, 95%CI [0.01, 0.02]), supporting Hypothesis 3(b).

The prevalence of adverse childhood experiences measured in this study was higher among the delinquent adolescents compared to other studies involving Chinese adolescents. For example, 89.4% of adolescents in a study based on 14,820 Chinese general adolescents reported one or more types of adverse childhood experiences (Wan et al., 2019). One possible explanation for this difference is the variation in subject groups between the two studies. The present study involved delinquent adolescents, while previous research has shown that juvenile violent offenders have experienced significantly more ACEs, including abuse, than non-offenders (Chang et al., 2021). Additionally, cultural differences in China must be considered to explain this variation (X. Li et al., 2014). On the one hand, “Seriousness and lack of warmth” is considered a worthwhile parenting style in China; on the other hand, many adolescents experienced parental absence and were cared for by non-parent guardians (usually grandparents), which increased their likelihood of exposure to ACEs (Zhou et al., 2020).

The study found that the status of adolescents experiencing childhood adversity predicted deviant peer affiliation, which means that experiencing adversity may have an impact on peer affiliation and friendship choices, specifically the more adversity adolescents experience the more likely they are to engage in deviant peer affiliation, which supports Hypothesis 1. This result was consistent with the results obtained by previous researchers in the general samples (Yoon et al., 2019; Jones et al., 2023). ACEs can lead to disruptions in multiple developmental processes from a developmental psychopathology perspective. These disruptions can result in impaired development in multiple domains of functioning, which may make adolescents less able to resist negative life influences such as deviant peer affiliation (Fanti et al., 2018; Morrow and Villodas, 2018).

This can also be explained by the early adolescent upbringing. In China, left-behind children lack warmth and protection from their families and face many mental health issues, forcing them to seek peer support and interaction (Zhao et al., 2015; Yang, 2022). The factor of adolescent role models has to be considered. Adolescents who typically experience more ACEs live in disadvantaged neighborhoods or schools and may face high rates of delinquent peer role models, which may exacerbate the likelihood that they will engage in deviant peer affiliation (Baglivio et al., 2017). Family dysfunction may weaken the family’s supervisory function and fail to protect adolescents from deviant peers or gangs. Prolonged exposure to violence may cause adolescents to misinterpret social norms, affecting the individual’s value construction and leading to more deviant peer affiliation (Trinidad, 2021). Overall, the results of this study suggest that ACEs predicted deviant peer affiliation.

The current study examined the mediating role of relative deprivation between adverse childhood experiences and deviant peer affiliation. The results confirmed Hypothesis 2. Specifically, ACEs may not only directly affect adolescents’ deviant peer affiliation, but may also lead to it by stimulating an individual’s sense of relative deprivation and thereby contributing to deviant peer affiliation. The results of this study also explain the phenomenon that not every individual who experiences childhood adversity engages in deviant peer affiliation. A possible explanation is that prolonged exposure to childhood adversity alters an individual’s perceptions (e.g., developing a sense of relative deprivation), leading to engagement in deviant peer affiliation. Scholars and clinicians believe that exposure to stressors (e.g., ACEs) early in the life course has direct effects on the immune, metabolic, and endocrine systems, and indirect effects through stress-related hormones that affect cognitive, social, and emotional development (Deighton et al., 2018; Jones et al., 2023). The theory of relative deprivation offers a rational explanation for this phenomenon. As adolescents’ peer affiliations expand and their individual cognition develops, they gradually recognize that they have experienced poverty and disadvantage in the past. Social comparisons then lead to a sense of unfairness or unmet needs. To rectify this, they may resort to shortcuts such as engaging with deviant peers (Yang et al., 2021; Li et al., 2023). In particular, researchers believe that the “envy effect” is widespread in China due to cultural and customary reasons, so that facing adversity or relative poverty leads to a stronger sense of relative deprivation, which in turn leads to transgressive behaviors or delinquency (Song et al., 2022).

In addition to the overall mediation model, each individual aspect of the mediation model deserves attention. Specifically, for the first component of the mediation process (i.e., ACEs→relative deprivation), the results of the present study indicated that ACEs were significantly and positively associated with relative deprivation. This is similar to previous findings such as that adolescents who reported parental neglect and estrangement believed that they deserved better treatment, and that individuals who had experienced childhood maltreatment report higher levels of individual relative deprivation (Rothman and Steil, 2012; Li et al., 2023). One possible explanation is that adolescents aspire to be treated as equals. However, through social comparisons, they may find themselves in a position of poverty and disadvantage, which can lead to a sense of relative deprivation. In the second part of the mediation process, this study found that relative deprivation was significantly and positively related to deviant peer affiliation. This finding is consistent with previous research indicating that individuals who experience greater relative deprivation are more likely to associate with deviant peers (Li et al., 2023). Self-determination theory may explain why individuals tend to satisfy their basic psychological needs. When negative experiences, such as relative deprivation, occur in real life, individuals may seek alternative ways to satisfy themselves, even if those ways are not recognized. For example, they may engage with deviant peers in order to gain higher peer status, more material wealth, or faster gratification (Lu et al., 2023). Simultaneously, research suggested that in adolescent antisocial peer groups, peer victimization, including aggression, is likely to be perceived as popular or “cool,” which may explain why adolescents choose deviant peer affiliation for self-gratification after experiencing relative deprivation (Chen et al., 2015).

Although studies have examined the effect of age differences on relative deprivation (Callan et al., 2015), few have focus on adolescents. Consistent with the hypothesis, the present study investigated the moderating role of age in the relationship between ACEs and relative deprivation among adolescents, thereby confirming that age moderates the indirect effects of ACEs on deviant peer affiliation through relative deprivation. Specifically, individuals who have experienced more ACEs in early adolescence tend to experience more relative deprivation. In contrast, this predictive effect diminishes in late adolescence. Childhood adversity, a toxic stressor, has been shown to impact brain development, cognitive performance, and self-regulation (McEwen and McEwen, 2017). Adolescents are in a developmental stage where they cope with stress and self-regulation. This may explain why adolescents at different ages differ in the impact of adverse childhood experiences and relative deprivation. Adolescents are not only individuals but also social beings. Changes in social comparisons can also affect individuals’ relative deprivation. During the initial stage of “separation-individualization,” early adolescents experience greater volatility compared to late adolescents who have a fixed peer group. This may be due to the fact that with the stabilization of social comparison factors and improved ability for cognitive self-regulation, the impact of ACEs on relative deprivation diminishes in late adolescence, and also predicts fewer deviant peer affiliations.

As with all studies, there are several important caveats to keep in mind for this study. Firstly, despite the theoretical support for this study and the extensive support from prior empirical research, we cannot establish causal relationships between variables due to the limitations of the cross-sectional research methodology. Future studies should validate these results using longitudinal tests or intervention experiments. Secondly, the results were derived from subjects’ self-reports, and due to the social approval effect, some subjects may have concealed their true situation, affecting the accuracy of the results. For example, not all participants may have been willing to disclose their ACEs. Future researches should use various experimental designs to mitigate the impact of the social approval effect. Thirdly, the data in this study were collected exclusively from Specialized Schools in China, and there is a lack of comparative data from students in ordinary schools. Therefore, the generalization of the results of this study needs to be approached with caution. This is because previous studies have shown that at-risk adolescents differ from normal adolescents in their cognitive and behavioral styles (Chang et al., 2021). Finally, due to gender differences in juvenile delinquency rates, most students in specialized schools are male, which may affect the results.

From a practical point of view, this study is instructive for protective services for delinquent youth in China. First, this study focused on a neglected group. Previous studies have primarily focused on delinquent and ordinary adolescents, but less attention has been given to minor delinquent adolescents. The results of this study can enrich the understanding of minor delinquent adolescents. Second, this study found that the reporting rate of ACEs among delinquent adolescents reached 97.4%, a grim figure that deserves our attention. On the one hand, we have to consider the cultural context of China, where harsh parental discipline, the challenging situation of left-behind children, and the neglect of children’s emotions all affect the healthy growth of adolescents. On the other hand, it also provides guidance for child protection services. This reminds authorities to prioritize the protective upbringing of children. For instance, when families are unable to provide a suitable environment for their children, protective schools, such as China’s specialized schools, can offer a better alternative. Third, it is necessary to offer adolescents who experience ACEs positive peer interaction models to enhance their resilience. For example, there is a need for specialized developmental programs to promote adolescents’ social skills, social competence and positive friendship relationships. Finally, this study emphasized the differential impact of age, reminding us to focus on the critical period of adolescent development, such as scientifically guiding adolescents to make reasonable cognitive attributions at the beginning of their secondary school years. For instance, psychological interventions and cognitive training can be used to reduce the negative effects of ACEs.