Oppositional defiant disorder (ODD), characterized by angry/irritable mood, argumentative/defiant behavior, and vindictiveness, is a prevalent developmental disorder worldwide with a prevalence of 3% to 5%, especially among Chinese school-age children and adolescents with a prevalence of 3.7% [1,2,3]. Children with ODD suffer from escalating risks of impairment in social, academic and occupational aspects, as well as other mental health problems in their lifespan [4, 5]. Among established multi-domain etiological factors (e.g., demographic, biological), family environment is considered pivotal in ODD development [2, 6]. Drawing on this view, Lin et al. [7], working from family system theory [8], proposed a multilevel family factors theory to understand family factors influencing ODD symptoms, categorizing them into three levels: entire (family characteristics), dyadic (family subsystems) and individual (family member) level [7]. Besides, these family factors at different levels were posited to be interrelated. However, while Lin et al. [7] provided a theoretical synthesis of these factors, empirical research systematically examining the longitudinal interrelationships across all three levels remains limited. Most existing studies have primarily examined factors at one or two levels, with few empirical studies investigating exactly how family factors at different levels are interrelated and associated with ODD symptoms.

From the perspective of developmental psychopathology, developmental cascade models offer an ideal framework to examine this process, where alterations in one domain dynamically influence and reinforce outcomes in others across time [9]. For instance, factors at the entire level might have cascading effects on ODD symptoms through affecting dyadic factors and individual factors. Adopting this model, Lavigne et al. [10] revealed the developmental cascade influences of multiple family factors on child ODD symptoms among a diverse community sample of 796 six-year-old children. Their findings demonstrated several significant cascading pathways: lower socioeconomic status predicted higher ODD symptoms via family conflict and parental scaffolding; parental depression predicted higher ODD symptoms via child negative affect; and parenting factors (hostility and support) influenced ODD symptoms via child effortful control.

Although this work provided important evidence for multidomain developmental cascades to ODD symptoms, two gaps remain. First, it is unclear whether similar cascade patterns operate among school-age children—a crucial developmental stage. School-age children experience transitions from simple, familiar environments to unfamiliar, complex environments, which may trigger ODD symptoms [11]. Second, Lavigne et al. [10] examined primarily unidirectional pathways from family factors to ODD symptoms. The influence in the opposite direction, however, has also been demonstrated [7, 12], necessitating investigation of bidirectional associations between ODD symptoms and family factors. Therefore, this study aims to examine the complicated associations between multilevel family factors and ODD symptoms by adopting a cross-level developmental cascade perspective among school-age children.

Factors at entire level associated with ODD symptoms

Entire level factors consist of family surface characteristics (e.g., social economic status, (SES)) and deep characteristics (e.g., family function). In this study, we focused on family SES as the entire level factor, as it is regarded as one of the most fundamental family background characteristics and has been a universal focus in child mental health research [13, 14]. Family SES captures a fundamental family context that characterizes the family’s access to social and economic resources [15]. According to the interactionist model [16], low SES is associated with inadequate education, a chaotic family environment, as well as reduced child-rearing investment and maladaptive parenting [16, 17]. These consequences would ultimately aggravate child externalizing problems, including ODD symptoms [16, 18]. Existing empirical research has consistently demonstrated low SES as a robust risk factor of ODD symptoms [10, 18,19,20]. For example, a cross-sectional study involving 622 preschool children from the general population indicated that low family SES was significantly associated with increased ODD symptoms [20]. Similarly, Lavigne et al. [10] found that family SES was significantly and negatively associated with child ODD symptoms one and two years later.

Factors at dyadic level associated with ODD symptoms

Dyadic level factors involve the functioning of each family subsystem, including marital subsystem and parent-child subsystem. In this study, we examined the conflict within the parent-child subsystem as the dyadic level factor. The parent-child subsystem is considered the most proximal to children’s daily experiences and directly shapes their emotional and behavioral socialization [8, 21, 22]. The coercive theory suggests that conflicts within this subsystem have been considered the most direct trigger of children’s oppositional and defiant behavior, as ODD symptoms are found to develop and maintain during the process of maladaptive parent-child interaction [23]. Moreover, the parent-child conflicts can arouse negative emotions in children, as well as affecting their emotional security and emotional socialization, which would exacerbate ODD symptoms [23]. Previous research has found that parent-child conflict plays a mediating role in the relationship between maltreatment and ODD symptoms [24]. Moreover, a recent network analysis including multiple family factors and ODD symptoms found that parent-child conflict, among various dyadic factors (e.g., interparental conflict, parenting style), was the most critical central and bridging influencing factor for ODD symptoms [21].

Factors at individual level associated with ODD symptoms

Individual level factors refer to the individual characteristics, cognitive factors, and emotion factors of each family member, including the parents and child. For parent factors, we focused on parental anger expression, as parental emotional expression serves as a key mechanism in child emotional socialization, with anger expression especially relevant to ODD symptoms [25, 26]. Anger expression refers to the way an individual processes and expresses anger. Here, we considered parental maladaptive anger expression (hereafter referred to as maladaptive anger expression unless otherwise specified), which has been shown to be related to child externalizing problems [25, 27, 28]. Specifically, parental anger expression usually triggers negative emotional experiences in their children, raising the level of their children’s trait anger and behavioral problems [25, 29]. Besides, when parents express their anger in a maladaptive way, children might internalize maladaptive ways of processing their emotions via observational learning, which in turn leads to their externalizing problems [28]. A recent empirical study involving 46 children diagnosed with ODD has revealed a significant positive association between parental anger expression and ODD symptoms [30].

For child factors, we focused on child self-control, a fundamental cognitive and behavioral regulatory capacity [31]. Self-control refers to the ability to control one’s emotions, thoughts, and behaviors to achieve desired outcomes [32] and has been consistently identified as crucial for behavior problems in children, including ODD symptoms [12, 33, 34]. According to the self-control theory [35], while good self-control enables individuals to effectively control and regulate negative affects [36], those with underdeveloped self-control are likely to experience difficulties in regulating negative emotions, consequently exhibiting more behavioral problems and delinquency [37]. Previous studies have shown that compared to typically developing children, ODD children have significant deficits in their ability to self-control [38], and this ability has also been demonstrated to be negatively associated with ODD symptoms [39]. Also, a recent study found that child self-control could predict ODD symptoms six months later, and emphasized the self-control deficits as a pivotal mechanism in the pathology of ODD [12].

The interrelationship of multilevel family factors and their bidirectional associations with ODD symptoms

According to the multilevel family factors theory [7], family factors at three levels are interrelated to influence ODD symptoms, with entire factors being most distal, dyadic factors proximal, and individual factors most proximal. Previous research has shown that entire factors (e.g., family SES) indirectly affect ODD symptoms through dyadic factors (e.g., parent-child conflict) and individual factors [10, 18, 20], and that dyadic factors (e.g., parent-child relationship) exert indirect effects through individual factors [40, 41]. Yet, most studies examined associations between two levels only [10, 18, 20, 40, 41], with few studies systematically investigating the interrelationships across all three levels. Moreover, most studies included only child individual factors [40, 41], leaving unclear how parent individual factors interrelated with other factors to influence ODD symptoms. Particularly, in the family system, parent individual factors are also posited to be distant to children, and potentially affect child ODD symptoms via child individual factors [8]. For example, maternal emotional socialization has been found to indirectly predict child behavioral problems through child self-control [42]. However, this pathway of parent individual to child individual to ODD has been neither theoretically conceptualized nor empirically examined. Given these limitations, the comprehensive interrelations among all three levels of family factors and their associations with ODD symptoms warrant further empirical clarification.

Furthermore, previous research suggests bidirectional associations between family factors and ODD symptoms. Children’s ODD symptoms can evoke negative parental emotions and escalating into frequent conflict within parent-child dyadic subsystem [43, 44]. Besides, parent individual factors of parental aggression and emotion dysregulation, have also been shown to escalate alongside children’s ODD symptoms over time, which can in turn influence these symptoms [43, 45]. In addition, child individual factors, such as child self-control, have also been proven to have a bidirectional connection with ODD symptoms [12]. Despite these bidirectional associations demonstrated, most existing studies have examined the bidirectional relationship between ODD and a certain single-level factor, lacking a comprehensive cross-level examination of these bidirectional associations.

The present study

Developing efficacious family-based intervention programs to prevent and mitigate ODD symptoms in school-age children necessitates an advanced understanding of how multilevel family factors are interrelated and associated with ODD symptoms. Thus, the present study examined a longitudinal development cascade model to investigate the developmental cascades from multilevel family factors to ODD symptoms and their bidirectional associations among school-age children (see Fig. 1). Investigating both is crucial because they, together, allow for a more comprehensive understanding of the dynamic nature of how ODD symptoms develop within the family environment. Specifically, as mothers are typically the main caregiver and interact more frequently with their children, we considered family SES at the entire level, mother-child conflict at the dyadic level, as well as maternal anger expression and child self-control at the individual level. We hypothesized that: (A) variables at each level not only had direct effects on ODD symptoms, but also had cascading effects on ODD symptoms via mediation by variables at a more proximal level. Specifically, (A1) family SES influences ODD symptoms via mother-child conflict, maternal anger expression and child self-control; (A2) mother-child conflict influences ODD symptoms via maternal anger expression and child self-control; (A3) maternal anger expression influences ODD symptoms via child self-control; (B) ODD symptoms have impacts on multilevel family factors, which in turn influenced subsequent ODD symptoms.

Fig. 1

Hypothesized developmental cascades for family SES, child ODD symptoms, mother-child conflict, maternal anger expression and child self-control