Intergenerational transmission of borderline personality disorder features, shame, and non-suicidal self-injury through perceived parental invalidation

Abstract

Background

While previous research suggests that borderline personality disorder (BPD) features may be transmitted from mothers to children through maladaptive parenting, little is known about the effect of the invalidating family environment, a critical etiological factor for BPD. Besides, there is a lack of research on fathers’ influence and the impact of parental BPD traits on adolescent shame and non-suicidal self-injury (NSSI). This study aimed to examine the mediating role of perceived parental invalidation in the intergenerational transmission of maternal and paternal BPD features on adolescent BPD features, shame, and NSSI using path analysis models.

Methods

Participants were a community sample of 142 father-adolescent dyads and 320 mother-adolescent dyads in China. Adolescents completed measures of their BPD features, perceived parental invalidation, sense of shame, and NSSI frequency, while parents reported their BPD features. Path analyses were conducted to investigate the direct and indirect effects of paternal and maternal BPD features on offspring outcomes via perceived parental invalidation, respectively.

Results

Perceived paternal invalidation was significantly associated with offspring BPD features, shame, and lifetime NSSI frequency. The effects of maternal BPD features on offspring BPD features and NSSI frequency were significant both directly and indirectly through perceived maternal invalidation as a mediator, but the relationship between maternal BPD features and offspring shame was fully mediated by maternal invalidation.

Conclusions

This study provides initial evidence that BPD features of mothers are transmitted to offspring BPD features, shame, and NSSI via perceived maternal invalidation, while there was no effect of fathers’ BPD features. The research and practical implications of the study were discussed.

Introduction

Borderline personality disorder (BPD) is a psychiatric illness characterized by affective instability, identity disturbance, behavioral impulsivity, and relationship impairment [1]. BPD is diagnosed in approximately 1–6% of the general population [1–3], among which many could be parents. It becomes not surprising that given the functional impairment, emotional dysregulation, and insecure attachment associated with BPD, the offspring of parents diagnosed with BPD or manifesting high levels of BPD features will face a heightened risk of adverse family environment and maladaptive parenting, leading to psychological maladjustment in various aspects [4, 5].

There exists established evidence that children of parents with BPD are associated with a wide range of poor psychopathological, social, and behavioral outcomes. First, familial aggregation of BPD is more of the norm than exception, with a 4- to 20-fold increase in the risk of BPD in relatives of BPD patients compared with the general population [6]. It is thus not surprising that parental BPD symptoms can predict the number of BPD symptoms in their offspring [7]. Moreover, the intergenerational transmission of psychopathology appears to be more generic than specific. In a longitudinal study, Auty and colleagues [8] found that the intergenerational transmission of personality disorder (PD) traits from fathers to daughters was both general and disorder-specific. Furthermore, two systematic reviews [4, 9] indicated a broad array of adverse offspring outcomes in terms of psychopathology (e.g., BPD traits, internalizing problems, externalizing problems), relationship quality (e.g., insecure attachment, poor relationships between the offspring and the parents with BPD), and psychological adjustment (e.g., emotional dysregulation, dysfunctional attitudes). The risk of psychopathology among the offspring of parents with BPD is even greater than the offspring of parents diagnosed with depression or other PDs [10, 11].

Empirical evidence has suggested two valid pathways of the intergenerational transmission of BPD. The first pathway is genetic heritability. The heritability of BPD is estimated to be 42% in a large cross-national twin sample [12], and a systematic review suggested that genetics can explain 42–70% of variance in the intergenerational transmission of BPD. More specifically, Fatimah et al. [13] analyzed the prevalence of BPD in twin families, adoptive families, and biological families, discovering that only the externalizing psychopathology of both parents and the BPD features of fathers constituted the genetic risk of offspring BPD features. However, if genetics is the only mediating factor, it cannot account for the fact that adoptive offspring of parents with BPD are also subjected to heightened risks of BPD and psychopathology in general.

Accumulating research has suggested that maladaptive parenting is a more pervasive and severe mediator in the intergenerational transmission of BPD. Compared to healthy controls, mothers with BPD report greater parental distress, emotional suppression, poorer parental satisfaction, and lower parenting efficacy, likely associated with the emotional and interpersonal impairments underlying the BPD psychopathology [14–17]. Indeed, parenting is inherently a stressful and emotion-provoking experience, thus requiring effective emotional regulation for appropriate parenting behaviors [18]. In this context, parents with BPD are more likely to engage in emotional, physical, and even sexual abuse [12, 19], resulting in trauma experiences among the offspring. Other maladaptive parenting behaviors have also been reported, including hostility, neglect [20], overprotection [14], fewer parental apologies [21], and less sensitivity and empathetic responsiveness [22, 23].

This dual-pathway model of the intergenerational transmission of BPD converges with one of the most widely accepted theories for the etiology of BPD, that is, Marsha Linehan’s biosocial theory [24]. It is proposed that the interaction of biological vulnerability to emotional reactivity and an invalidating family environment posits the risk of developing and maintaining BPD psychopathology [25, 26]. In an invalidating family environment, children’s emotional and personal experiences are treated with parental neglect, denial, and punishment, resulting in a series of dysfunctions including difficulties in identifying and regulating emotions, a lack of self-concept clarity, and interpersonal impairments [24]. Cross-cultural studies have also suggested cultural specificity in the expression of parental invalidation. For example, caregivers from a Chinese invalidating family environment are more likely to engage in the neglect and denial of emotional experiences, overemphasis of academic achievements, and psychological control [27].

It has been vastly demonstrated that the invalidating family environment is a risk factor for not only BPD but various internalizing and externalizing symptoms [28], as well as other forms of maladjustment closely associated with BPD, such as the sense of shame [29] and self-injurious behaviors [30, 31]. Moreover, recent studies have suggested that parental invalidation is the key mediator in the intergenerational transmission of BPD pathology. Buckholdt and colleagues [28] reported that parents experiencing heightened levels of emotional dysregulation, a core deficit in BPD, were more likely to be perceived as invalidating their offspring’s affect, which in turn was associated with higher levels of offspring emotional dysregulation.

While maladaptive parenting and parent-child conflicts in general are repeatedly identified as significant mediators in the intergenerational transmission of BPD [4, 32], several literature gaps remain under-investigated. First, the majority of existing research focused on mothers with BPD only. This is not surprising given the disproportionate sex distribution of BPD, as approximately 75% of BPD diagnoses are seen among females [1, 2]. However, this statistic was based on the clinical population, as recent reviews revealed comparable distribution between sexes in community samples [33, 34]. Auty and colleagues [8] also found both generic and disorder-specific intergenerational transmission of PD traits in fathers. Besides, in a twin and adoptive study, Fatimah et al. [13] found that only father’s BPD contributed to the genetic risk of BPD in children, not mother’s BPD, suggesting gender-specific mechanisms of BPD intergenerational transmission. Second, while most previous studies focused on offspring BPD and general psychopathology, the effect of parental BPD on other maladjustments related to BPD remains unclear. For example, shame is not only a core affect but also the driver of various maladaptive behaviors in BPD pathology [35]. Parental invalidation also plays a significant role in shame-proneness, as individuals are more likely to develop the tendency of self-invalidation and thus feel shameful about their negative affect [29]. While Macfie and Swan [36] reported that children of mothers with BPD were more likely to have shame self-representations (e.g., I am bad), this study used a qualitative design instead of standardized self-report measure of shame experience. Furthermore, individuals subjected to a heightened sense of shame posit great risks of engaging in self-injurious behaviors to relieve negative emotions [37]. In fact, non-suicidal self-injury (NSSI), defined as deliberate harm to body tissue with no suicidal ideation [38], is a maladaptive behavior closely intertwined with BPD psychopathology. NSSI is listed as a diagnostic criterion for BPD in DSM-5 [1], and the cooccurrence of NSSI and BPD is associated with higher severity and comorbidity with other psychiatric disorders [39]. There is also a robust association between parental invalidation and adolescent NSSI, as individuals growing up in an invalidating family are more likely to engage in NSSI as a maladaptive practice of emotional regulation [31]. Finally, most theoretical conceptualizations and empirical evidence of the intergenerational transmission of BPD were built upon Western cultures. The effect of maladaptive parenting behaviors may be attenuated when they are culturally normative [40, 41]. Given the traditional value of prioritizing accomplishments over emotional well-being in China, despite the potentially greater prevalence of parental invalidation, its effect may be less salient on Chinese children. Even though emerging evidence has demonstrated a cross-cultural intergenerational impact of parental rejection on offspring psychopathology [42], there remains a need to understand whether invalidation, as a specific form of maladaptive parenting behavior, is a cross-cultural transmission mechanism for BPD psychopathology.

The current study

In response to the above literature gap, the current study aims to understand the invalidating parenting behavior and outcomes of children of mothers and fathers with BPD features using a large cohort of Chinese adolescents. The hypothetical model is shown in Fig. 1. We expected that higher parental BPD features would be associated with poorer offspring outcomes, including higher BPD features (path e), heightened sense of shame (path f), and more frequent NSSI (path g), both directly and indirectly through perceived parental invalidation (paths a and b/c/d). Additionally, adolescent BPD features are anticipated to be positively associated with shame and NSSI. In order to address the lack of paternal data in existing literature, our study analyzed the effects of maternal and paternal BPD features in separate structural equation models. It is also worth noting that as our data are cross-sectional, the direction of paths remains conceptual and cannot infer causality.

Fig. 1.

Conceptual path model

Methods

Participants and procedure

We first recruited 3007 families, each including one adolescent and at least one parent (62 families with both parents participating) from middle and high schools in Fujian, China. To ensure response validity, we excluded 962 parent-adolescent dyads who failed the verification questions and 162 who did not complete the questionnaires. Only parent-adolescent dyads in which the parent reported at least one BPD feature (obtaining a total score of at least 1 on the McLean Screening Instrument for Borderline Personality Disorder) were included in the final analysis, resulting in a final sample of 455 adolescents (M_age = 14.33, SD_age = 1.38; 55.4% female; 98.9% Han ethnicity; 65.3% had experienced being left behind, meaning that one or both parents had been absent for a continuous period of at least three months due to labor migration) and 462 parents. Among the adolescents, 135 participated with fathers, 313 participated with mothers, and 7 participated with both parents. The demographics of the parents are presented in Table 1. All adolescent and parent participants completed the scales on an online platform following standardized administration procedures after informed consent was obtained. Ethical approval was granted by Peking University (Project Number: 2022-02-06).

Table 1.

Demographic information of the parents

	Father (n = 142)	Mother (n = 320)
Age in years (M ± SD)	44.28 ± 5.56	40.40 ± 5.23
Annual overall household income (%)
Less than ¥10,000	12.0	9.1
¥10,000–50,000	31.7	39.7
¥50,000–100,000	30.3	30.9
¥100,000–200,000	18.3	17.2
More than ¥200,000	7.7	3.1
Education level (%)
Junior high school or below	66.2	68.8
Vocational school	13.4	13.1
Senior high school	9.9	6.6
Associate degree	5.6	7.8
Bachelor’s degree or above	4.9	3.8

Measures

McLean screening instrument for borderline personality disorder (MSI-BPD)

The MSI-BPD is a self-report screening tool for BPD based on the DSM-IV clinical interview [43]. It consists of 10 yes-or-no items corresponding to the nine DSM-IV diagnostic criteria of BPD (the ninth criterion, paranoia and dissociation, is assessed by two items), with total scores ranging from 0 to 10. Higher total scores indicate greater levels of BPD features. In this study, the Chinese version of MSI-BPD was administered to both adolescents and parents to assess their BPD features, demonstrating good internal consistency for adolescents (α = 0.83), fathers (α = 0.75), and mothers (α = 0.82).

Chinese invalidating family environment scale (CIFS)

The first part of CIFS [44] was administered to adolescents to measure the level of perceived paternal and maternal invalidation. CIFS is a retrospective self-report questionnaire to assess the invalidating family environment in Chinese cultural contexts. It consists of two subscales that respectively measure the extent of invalidating parenting of the father with 20 items and the mother with 27 items. Each item is scored on a 4-point scale (1 = never; 4 = always). Higher total scores indicate greater levels of perceived parental invalidation. In the present study, the internal consistency was good for both subscales (father: α = 0.91; mother: α = 0.95).

Experience of shame scale (ESS)

Adolescents’ shame was assessed using the ESS [45]. The ESS is a 25-item self-report measure of the frequency of shame experiences over the past year across three clinically relevant domains: characterological, behavioral, and bodily shame. Each item is scored on a 4-point scale (1 = not at all; 4 = very much), with higher total scores indicating greater shame. Cronbach’s α for ESS in the current sample was 0.96.

Inventory of statements about Self-Injury (ISAS)

Adolescent participants were first asked whether they had ever engaged in NSSI. Those who answered “yes” were administered the behavioral section of the ISAS, which assessed lifetime frequency of 12 NSSI behaviors [46], including banging or hitting oneself, biting, burning, carving, cutting, wound picking, needle-sticking, pinching, hair pulling, rubbing skin against rough surfaces, severe scratching, and swallowing chemicals. Participants were asked to estimate how many times they had intentionally performed each behavior without lethal intent. The ISAS has demonstrated good reliability and validity in previous research [47]. Following Hamza and Willoughby’s procedure [48], we summed the frequency of the 12 behaviors and recoded the total into seven ordered categories: no incident, 1 incident, 2–4 incidents, 5–10 incidents, 11–50 incidents, 51–100 incidents, and more than 100 incidents.

Data analysis

Descriptive statistics and correlations of the variables of interest were calculated using SPSS 30 [49]. Pearson correlation coefficients were used for continuous variables, while Spearman correlation coefficients were used for lifetime NSSI frequency (an ordinal variable). Path analysis was performed for father-adolescent and mother-adolescent dyads using the Lavaan package [50] in R. Given that lifetime NSSI frequency was an ordinal variable, we used diagonally weighted least squares estimation to estimate the model parameters, which also accommodated the non-normal distribution of other model variables. Preliminary analyses included adolescents’ gender, age, overall household income, and parental education level as covariates in the path analysis models. Results indicated that only adolescents’ gender significantly predicted outcome variables and was therefore retained in the final models. All p values were two-tailed, and findings at the p <.05 level were considered significant.

Results

Descriptives and correlations

Descriptive statistics and correlations among study variables of father-adolescent and mother-adolescent dyads are presented in Tables 2 and 3. Among the 455 adolescents, 80.7% reported no NSSI behavior, 0.7% had performed NSSI once, 2.6% had performed NSSI 2–4 times, 1.8% had performed NSSI 5–10 times, 7.3% had performed NSSI 11–50 times, 2.0% had performed NSSI 51–100 times, and 5.1% had performed NSSI more than 100 times. For father–child dyads, paternal BPD features were not significantly correlated with any of the other variables of interest, but all other variables were significantly positively correlated with each other. For mother-adolescent dyads, all variables of interest were significantly positively correlated with each other.

Table 2.

Descriptive statistics and correlations among study variables in father-adolescent dyads (n = 142)

	M	SD	1	2	3	4
1. Paternal BPD features	2.23	1.62	-
2. Paternal invalidation	38.88	11.31	0.08	-
3. Adolescent BPD features	2.70	2.64	0.12	0.34***	-
4. Adolescent shame	55.35	15.42	− 0.02	0.30***	0.46***	-
5. Adolescent lifetime NSSI frequency	-	-	0.11	0.25**	0.48***	0.26**

The correlation coefficients between NSSI frequency and other variables are Spearman correlation coefficients, and the coefficients among the other variables are Pearson correlation coefficients. **p <.01, ***p <.001

Table 3.

Descriptive statistics and correlations among study variables in mother-adolescent dyads (n = 320)

	M	SD	1	2	3	4
1. Maternal BPD features	2.67	2.03	-
2. Maternal invalidation	55.30	16.62	0.20***	-
3. Adolescent BPD features	3.12	2.83	0.33***	0.44***	-
4. Adolescent shame	53.83	16.11	0.16**	0.41***	0.51***	-
5. Adolescent lifetime NSSI frequency	-	-	0.21***	0.27***	0.50***	0.26***

The correlation coefficients between NSSI frequency and other variables are Spearman correlation coefficients, and the coefficients among the other variables are Pearson correlation coefficients. **p <.01, ***p <.001

Path analysis

Path model of paternal BPD features

Model 1 examined the intergenerational transmission of paternal BPD features via perceived paternal invalidation (see Fig. 2). Results showed that paternal BPD features were not significantly associated with adolescent-reported paternal invalidation (b = 0.73, SE = 0.47, p =.115, β = 0.11; path a). However, paternal invalidation positively predicted adolescent BPD features (b = 0.08, SE = 0.02, p <.001, β = 0.36; path b), adolescent shame (b = 0.47, SE = 0.10, p <.001, β = 0.34; path c), and adolescent lifetime NSSI frequency (b = 0.03, SE = 0.01, p <.001, β = 0.34; path d). The direct effects of paternal BPD features on adolescent BPD features (b = 0.10, SE = 0.12, p =.377, β = 0.06; path e), shame (b = −0.88, SE = 0.85, p =.299, β = − 0.09; path f), and NSSI frequency (b = 0.07, SE = 0.08, p =.397, β = 0.11; path g) were not significant, and the indirect effects of perceived paternal invalidation mediating the relationship between paternal BPD features and adolescent BPD features (b = 0.06, SE = 0.04, p =.139, β = 0.04), shame (b = 0.34, SE = 0.23, p =.139, β = 0.04), and NSSI frequency (b = 0.02, SE = 0.02, p =.156, β = 0.04) were not significant either.

Fig. 2.

Standardized parameter estimates for model 1. IND indirect effects; *p <.05, ***p <.001.

Path model of maternal BPD features

Model 2 investigated the intergenerational transmission of maternal BPD features via perceived maternal invalidation (see Fig. 3). Maternal BPD features positively predicted adolescent-reported maternal invalidation (b = 1.65, SE = 0.46, p <.001, β = 0.20; path a). Maternal invalidation, in turn, positively predicted adolescent BPD features (b = 0.07, SE = 0.01, p <.001, β = 0.39; path b), adolescent shame (b = 0.38, SE = 0.05, p <.001, β = 0.39; path c), and adolescent lifetime NSSI frequency (b = 0.02, SE = 0.00, p <.001, β = 0.34; path d). We examined the indirect effects to further test whether perceived maternal invalidation mediated the relationship between maternal BPD features and adolescent outcomes. Results showed that the indirect effects of maternal invalidation mediating maternal BPD features and adolescent BPD features (b = 0.11, SE = 0.03, p =.001, β = 0.08), shame (b = 0.63, SE = 0.20, p =.001, β = 0.08), and NSSI frequency (b = 0.04, SE = 0.01, p =.002, β = 0.07) were all significant. Additionally, the direct effects of maternal BPD features on adolescent BPD features (b = 0.33, SE = 0.07, p <.001, β = 0.24; path e) and NSSI frequency (b = 0.08, SE = 0.04, p =.028, β = 0.16; path g) remained significant after accounting for the mediation. However, the direct effect of maternal BPD features on shame (b = 0.52, SE = 0.44, p =.239, β = 0.07; path f) was not significant, suggesting that the relationship between maternal BPD features and offspring shame was fully mediated by perceived maternal invalidation.

Fig. 3.

Standardized parameter estimates for model 2. IND indirect effects; *p <.05, **p <.01, ***p <.001.

Discussion

The goal of our study was to investigate the intergenerational transmission of maternal and paternal BPD features on offspring BPD features and BPD-related psychological maladjustment, including shame and NSSI frequency, and whether this transmission can be explained by perceived family invalidation, a key etiological factor for BPD. To the authors’ knowledge, this is the first empirical study investigating the effects of both maternal and paternal BPD features on offspring shame and NSSI in a large Chinese community sample using a culturally sensitive measure of invalidating parenting. We anticipated that perceived parental invalidation would mediate the intergenerational transmission of parental BPD features to offspring BPD features, shame, and NSSI, while the outcome variables would mutually correlate. Although specific hypotheses for the pathways of fathers and mothers were not proposed due to the lack of former literature on the effects of paternal BPD features, results differed for the paternal and maternal models, suggesting different mechanisms for their intergenerational transmission.

Our findings indicated that for mothers, through perceived maternal invalidation, BPD features significantly influenced offspring BPD features and NSSI frequency both directly and indirectly, and significantly influenced offspring shame indirectly. These results largely replicated and extended previous evidence demonstrating that offspring of mothers with BPD are at greater risk of developing BPD features and experiencing various emotional and behavioral challenges [4, 36]. We also found that perceived maternal invalidation substantially mediated the relationship between mothers’ BPD features and the maladaptive outcomes in offspring. While theoretical models have supported the invalidating family environment as a key etiological factor for the development and maintenance of BPD [24, 25], the role of parental invalidation in the intergenerational transmission of BPD remains unclear [32]. In fact, mothers with BPD may also have grown up in an invalidating family environment when they were children, thus internalizing and normalizing the invalidating parenting behaviors in their practice. This proposition was supported by evidence that not only negative parenting behaviors in general [41, 51, 52], but parental invalidation specifically [53], is transmitted across generations. Our findings suggested that for BPD features and NSSI frequency, perceived maternal invalidation only partially mediated the transmission of BPD features. As mothers are typically the primary caregivers and thereby interact with the children in various contexts, it is possible that other maternal dysfunctions and negative patterns of parenting, such as emotional suppression and abuse [16, 19, 54], also play important roles in the intergenerational transmission of BPD traits. Besides, we only examined the family environment component of the biosocial model, leaving the degree of biological influence unknown. This may also explain why shame was fully mediated by perceived maternal invalidation, as shame is a self-conscious emotion with low to moderate genetic heritability and a substantial non-shared environment component [55].

In contrast, the path model for fathers found that paternal BPD features had no direct or indirect effects on offspring psychopathology outcomes and were not associated with perceived paternal invalidation, while paternal invalidation was significantly associated with more severe BPD features, shame, and NSSI. These findings suggested that, unlike hypothesized, paternal BPD features were not transmitted to offspring maladaptive outcomes. However, in Fatimah et al. [13], paternal BPD traits significantly predicted child BPD traits in only biological offspring but not adoptive offspring, implying genetic transmission of BPD traits. One possible explanation may be that, due to the nonclinical nature of our sample, there was a robust floor effect of BPD features, not allowing the manifestation of genetic heritability. It was also surprising that, unlike mothers, fathers with higher BPD features were not more likely to be reported by their children as invalidating. The majority of our sample (65.3%) reported experiencing being left behind, a prevalent phenomenon in which children stay in their original residence while one or both parents migrate to earn more money. The high rate of paternal absence is a notably common pattern in Chinese left-behind families as a result of the traditional Confucian gender norms, which often assign men the role of economic providers [56, 57]. Both the influence of paternal BPD traits on children’s perceived paternal invalidation and their intergenerational transmission may be attenuated by the limited father-child interactions. Besides, we measured the level of invalidating family environment holistically. As previous literature stressed predominantly genetic transmission of fathers’ BPD features, potentially, only paternal emotional dysregulation was playing a role in the process, not the other behavioral invalidation. Given the gender and cultural specificity, it is worth exploring whether the subdomains of parental invalidation mediate BPD transmission differently. In addition to the above explanations, the interaction between maternal and paternal BPD features remained untested in our study due to the lack of available information. It is noted that individuals diagnosed with BPD are more likely to engage in relationships with other people with BPD, and their relationships are characterized by higher frequency and intensity of miscommunication and violence [58]. In this case, future research is recommended to examine maternal and paternal BPD traits simultaneously and their mutual impact on children’s development.

This study contains several limitations. Firstly, the design of the current study is cross-sectional in nature, so we were unable to examine the causal relationships between parental BPD features, perceived parental invalidation, and adolescent BPD features, shame, and NSSI. Longitudinal research is needed to further clarify the direction of the hypothesized causality. Another limitation is the representativeness of the sample, which consisted of nonclinical Chinese adolescents and their parents. It is possible that adolescent inpatients or outpatients, and adolescents of other races or countries may report BPD traits, perceived parental invalidation, feelings of shame, and NSSI experiences differently. Future studies can conduct similar surveys and analyses in clinical samples as well as adolescents and their parents from diverse cultural backgrounds to replicate our findings. Additionally, we mainly collected data from one of the parents, of which the majority were still mothers. Future studies could target participation from both parents to examine the difference in the influence of parental BPD and invalidation on adolescent psychopathology as well as the potential interaction between the two parents.

Finally, there were limitations with the instruments used in this study. As MSI-BPD is a binary self-report measure, it is well-suited for screening purposes but less effective in evaluating the severity of BPD features. For example, the Borderline Personality Features Scale is a psychometrically sound self-report instrument for assessing BPD features in non-clinical youth [59], and could be used in future studies to obtain a more accurate assessment of BPD severity. For parental invalidation, while self-report measures can evaluate subjective perceptions, many studies have found that BPD features are associated with biased response styles, including impulsivity, dichotomy, and heightened sensitivity [60–62]. More methods of assessment, such as standardized interviews on BPD features and NSSI, ecologically valid observations of family interactions, and implicit association test of shame, are recommended to be included in future research [63].

Notwithstanding the above limitations, the results from this study contribute to the literature. Firstly, our sample was Chinese adolescents, and this is vital as few studies on the intergenerational transmission of BPD features were conducted with adolescents from non-Western cultures. In fact, a notable difference from studies based on the Western population [13] is that paternal BPD features were not associated with paternal invalidation or offspring outcomes. Given the high prevalence of being left behind in our sample and mothers being mainly responsible for Chinese families, interaction time would be a significant moderator to study parenting behaviors in Chinese contexts. Second, we examined BPD features of both fathers and mothers, while most existing studies only collected maternal data. Findings differed between parents, addressing the need to identify the potentially different parental influences on offspring’s BPD features, shame and NSSI. Future research should explore the specific mechanisms of paternal influence and consider the moderating role of cultural variables. Besides, results of our study provided preliminary supporting evidence for the mediating role of parental invalidation in BPD transmission, which can benefit the replication and expansion on other mediation pathways (e.g., shame in the relationship between invalidation and NSSI) with a larger longitudinal sample. Finally, this study justifies the necessity of investigating the risk of biological vulnerability and parental invalidation in combination, potentially by recruiting twin and adoptive families to disassemble genetic and environmental influence. Our results are mostly consistent with previous studies illustrating that mothers’ BPD features can be transmitted to adolescents via parental invalidation [25, 26, 28], and we extend these findings by demonstrating the impact of parental BPD features on shame and NSSI.

Besides, our study could have several meaningful applied implications, especially if the results can be replicated among clinical samples. For instance, teaching the interpersonal validation-oriented skills from dialectical behavior therapy in addition to parenting and adolescent emotion regulation skills in intervention programs can be beneficial for fostering parent-child relationships and thus addressing adolescent emotional and behavioral dysregulation [64, 65]. Furthermore, despite the non-significant findings related to paternal BPD features, the effects of perceived paternal invalidation remained robust, consistent with a previous study in which fathers played an important role in the development of BPD in adolescents [13]. Therefore, fathers should be included and encouraged to participate more actively in the intervention of adolescent BPD. Lastly, shame and NSSI are two distinct features closely related to BPD in adolescence [29, 31]. Both shame and NSSI should be given more attention and taken as essential targets in the early prevention and intervention of adolescent BPD.

Conclusions

Despite limitations in sample diversity and methodology, our study makes a significant contribution to the current understanding of the intergenerational transmission of BPD features. We investigated maternal and paternal effects separately and did find that they have different mechanisms. Specifically, mothers’ BPD features influenced offspring maladjustments directly and indirectly through maternal invalidation, while fathers’ BPD features were not transmitted to offspring psychopathology. For mothers, these associations applied to not only offspring BPD features but also shame experience and NSSI frequency. While implications of these findings require further investigation on the biological underpinnings and interparent dynamics, the present study highlighted that both sides of caregivers should be the target of validation interventions for at-risk adolescents.

Abbreviations

BPD

Borderline Personality Disorder

NSSI

Non-Suicidal Self-Injury

PD

Personality Disorder

MSI-BPD

McLean Screening Instrument for Borderline Personality Disorder

CIFS

Chinese Invalidating Family Environment Scale

ESS

Experience of Shame Scale

ISAS

Inventory of Statements about Self-Injury

IND

Indirect Effect

Authors' contributions

ZL: Conceptualization, Investigation, Methodology, Writing - original draft preparation, Writing - review and editing; YC: Data curation, Formal analysis, Investigation, Visualization, Writing - original draft preparation, Writing - review and editing; XZ: Funding acquisition, Resources, Supervision, Writing - review and editing; QW: Conceptualization, Investigation, Project administration, Supervision, Writing - original draft preparation, Writing - review and editing. All authors read and approved the final manuscript.

Funding

This study was supported by Shanghai Education Science Research Program: C2025171.

Data availability

Data and materials are available upon request from the corresponding author.

Declarations

Ethics approval and consent to participate

This research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. Ethical approval was granted by the Ethical Committee of Peking University (project reference number: #2022-02-06). Written informed consent was obtained from all participants and their parents.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.