1 Introduction

Pathological narcissism involves distorted and/or maladaptive regulation of self-image, such that affected individuals inordinately rely on admiration, exploitation, self-concealment, or grandiose fantasies to feel acceptable and adjusted [1, 2]. Indeed, pathological narcissism often involves self-elevating attitudes and behaviours (narcissistic grandiosity) intermingled or fluctuating with fragile self-esteem and hypersensitivity (narcissistic vulnerability) [3]. Pathological narcissism is associated with interpersonal and psychological distress [4], maladaptive emotion regulation [5], and aggression proneness [6]. Given such difficulties, which negatively affect not only the individual but others around them [7], it is important to understand associated personality factors that may underlie pathological narcissism to inform intervention principles, especially since empirically-supported treatments for this condition are lacking [8, 9].

From the perspective of the five-factor trait model of personality, pathological narcissism has been strongly linked with antagonism, reflecting high disagreeableness, along with neuroticism and extraversion, with the latter two components showing different associations with vulnerability and grandiosity [10]. However, measures of clinically-salient narcissistic personality, representing both grandiose and vulnerable features, were linked with all three of these five-factor domains. Similarly, pathological narcissism comprising both domains, measured with the Pathological Narcissism Inventory (PNI) [11], has shown consistent positive associations with neuroticism and negative associations with agreeableness across community and clinical samples [12].

Beyond five-factor models of personality, psychodynamic theories have emphasized compromised internal relational representations––outlined by attachment theory––and impaired personal identity as fundamental to narcissistic pathology [13, 14]. These perspectives suggest that pathological narcissism may be characterized by a relative lack of integrated internal models of self-in-relation-with-others, involving diminished trust and connectedness in relationships and unstable self-concept. Research regarding these constructs separately has shown pathological narcissism to be associated with aspects of attachment insecurity [12, 15], particularly attachment anxiety [16], and dysfunctional identity, characterized by difficulties maintaining a sense of consistency in personal values, beliefs, and self-concept across time and contexts [15, 17, 18]. However, it is unclear whether attachment and identity domains together account for pathological narcissism, and whether they do so beyond the effects of five-factor traits. Furthermore, given that attachment and identity are influenced developmentally by childhood interactions with parental figures, it is plausible that these domains could mediate the known effects of adverse parenting on pathological narcissism [19]. In other words, attachment insecurity and identity dysfunction might account for the relationship between adverse parenting and pathological narcissism.

The present study investigated insecure attachment and dysfunctional identity as factors in pathological narcissism beyond the effects of five-factor traits. Determining correlates and potential contributing variables such as attachment and identity can help focus interventions for pathological narcissism. In clinical presentations and contexts, grandiosity and vulnerability are typically intertwined; hence, in keeping with previous research on clinically significant narcissistic dysfunction [4, 12, 20], we examined overall pathological narcissism as a criterion variable. Insecure attachment––comprising anxiety and avoidance––and dysfunctional identity were each hypothesized to be linked with pathological narcissism after controlling for relevant five-factor personality traits. Given theorized origins of attachment and identity in childhood relational experience, an exploratory mediation model was also tested whereby attachment insecurity and dysfunctional identity would mediate the relation between maladaptive parenting and pathological narcissism.

2 Methods

Participants were 392 adults sourced online through Prolific Academic (https://www.prolific.com/). Participants were informed and provided consent by completing study questionnaires using Qualtrics. Remuneration of £6 was provided to participants. A total of 401 individuals, mostly from the UK (80%), consented to participate. Removal of 9 cases due to inattentive responding (i.e., failed attention checks) and incompleteness yielded a final sample of N = 392.

Participants’ average age was 34.30 (SD = 12.66; range = 18–75); 69.4% identified as female, 29.1% as male, and 1.3% as non-binary. Most, 78.6%, identified as Caucasian, 7.7% as Asian, 3.6% as South Asian, and 10% as other ethnicities.

An abbreviated version of the Pathological Narcissism Inventory, the Brief PNI (B-PNI) [21] was used to assess pathological narcissism. The B-PNI comprises narcissistic grandiosity and vulnerability, which are clinically observed to overlap [1, 2] and were highly correlated in this sample (r = 0.69, p < 0.001); the mean of all 28 items provided a total score (α = 0.92). Five-factor personality was assessed using the 20-item International Personality Item Pool (mini-IPIP) [22], comprising neuroticism (α = 0.71), extraversion (α = 0.80), openness (α = 0.65), agreeableness (α = 0.80), and conscientiousness (α = 0.73). Attachment insecurity was assessed using the 12-item short form of the Experiences in Close Relationships Scale (ECR-SF) [23], comprising attachment anxiety (α = 0.78) and attachment avoidance (α = 0.80). Dysfunctional identity was measured with the Self-Concept and Identity Measure (SCIM) [24], using the total of 27 items to reflect overall impairment in personal identity (α = 0.91). Finally, a composite of maternal and paternal scores from the Measure of Parental Style (MOPS) [25] reflected perceptions of adverse interactions with parents during childhood (α = 0.93).

Analyses, conducted with SPSS 27 and PROCESS 3.5 [26], included zero-order correlations, including checking age and gender as covariates, followed by linear regression predicting pathological narcissism. Significant five-factor traits, insecure attachment domains, and dysfunctional identity were entered in separate steps. Subsequent regression tested indirect effects of adverse parenting on pathological narcissism through attachment insecurity and dysfunctional identity as mediators, using bootstrapped 95% confidence intervals.

3 Results

Examination of age and gender as potential covariates revealed age to be significantly negatively associated with pathological narcissism, r = − 0.35, p < 0.001, in that younger individuals reported higher levels of pathological narcissism. No significant association was found for gender. Table 1 presents descriptive statistics and zero-order correlations for main study variables. Pathological narcissism was positively and moderately associated with neuroticism and moderately and negatively with conscientiousness, though not with other five-factor personality traits. Attachment anxiety and attachment avoidance were positively associated with pathological narcissism, as was dysfunctional identity. Thus, individuals higher in pathological narcissism endorsed tendencies toward more unstable emotion and less carefulness and self-direction, along with reduced self-concept and relational security.

Table 1 Descriptive statistics and zero-order correlations among main study variables, N = 392
Full size table

A regression model subsequently examined those variables which were significant at the level of zero-order correlations as predictors of pathological narcissism. Variables were entered in separate steps for age, neuroticism and conscientiousness, attachment anxiety and avoidance, and finally dysfunctional identity. The overall regression model was significant, F(6385) = 62.68, p < 0.001, with only age, attachment anxiety, and dysfunctional identity significant in the final model accounting for all predictor variables (Table 2). In other words, after controlling for age and significant five-factor traits, insecure attachment––mainly attachment anxiety––and identity dysfunction were prominent factors in pathological narcissism, accounting for 17% and 8% of the variance, respectively. These significant predictor variables were subsequently examined as parallel mediators in a path model of the indirect effect of adverse parenting on pathological narcissism.

Table 2 Standardized coefficients from regression analyses examining attachment insecurity and dysfunctional identity predicting pathological narcissism, after accounting for age and five-factor personality variables
Full size table

Figure 1 indicates the hypothesized mediation model including standardized path coefficients and estimates of indirect effects. Since bootstrapped confidence intervals for each mediator did not contain zero, significant mediation effects were observed for both attachment anxiety and dysfunctional identity. Perceived adverse parenting was no longer significant with mediators included, indicating full mediation. Thus, both attachment anxiety and dysfunctional identity helped explain the relation between adverse parenting and pathological narcissism, in that a 1 SD increase in adverse parenting corresponded to a 0.2 SD increase in pathological narcissism through the combined effect of attachment anxiety and identity dysfunction.

Fig. 1
figure 1

Standardized coefficients and indirect effects for attachment anxiety and dysfunctional identity as mediators of the relationship between adverse parenting and pathological narcissism

Full size image

4 Discussion

Findings from this preliminary investigation indicate that both attachment anxiety and dysfunctional identity account for a significant degree of narcissistic pathology beyond the effects of age and five-factor traits, which were non-significant when examined with the inclusion of attachment and identity variables. The present findings indicate that attachment anxiety and impaired identity––reflecting unstable representations of self-other relations and inconsistent self-concept––may be particularly salient to understanding the form of narcissistic dysfunction captured by the PNI, characterized by contingent self-esteem, self-concealment, and retreat into grandiose fantasy. These features of pathological narcissism thus seem entwined with confusion about one’s sense of self coupled with insecurity about the self in relation to others, involving fears of rejection and abandonment. While previously examined separately [16, 17], attachment anxiety and identity dysfunction appear to be simultaneously relevant to understanding pathological narcissism, and perhaps more so than five-factor traits, including tendencies toward negative emotionality and lack of conscientiousness. These findings also raise questions about the suggestion, based on convergent validity research, that identity dysfunction may largely reflect maladaptive neuroticism [27]. Moreover, problems with both attachment and identity accounted for the relationship between adverse parenting and pathological narcissism, in that individuals who characterized their parents’ behaviour toward them as maladaptive had more anxious attachment and identity dysfunction, which in turn linked with higher pathological narcissism.

The present findings align with contemporary psychodynamic theories regarding contributory factors to pathological narcissism [14, 28]. These perspectives consider early difficulties in caregiver responsiveness to instil patterns of insecure relatedness that interfere with relatively normal identity achievement, producing an unstable self-concept that the individual attempts to regulate using narcissistic mechanisms such as self-enhancing fantasies (grandiosity) and self-concealment (vulnerability). Aspects of attachment and identity may thus be important targets for clinical intervention in pathological narcissism, as suggested by psychodynamic treatment models [29], including exploration of potentially traumatic developmental experiences. While further research extending the present findings may help to inform treatment approaches, more studies directly examining the treatment of pathological narcissism are urgently needed [2], particularly since many clinicians are likely encountering patients with high levels of narcissistic dysfunction in their practices [30]. Indeed, recent psychotherapy process research among patients with narcissistic pathology suggests promising directions for treatment in this population [31, 32].

Interestingly, associations typically observed between five-factor traits and pathological narcissism (e.g., agreeableness/antagonism) [10, 12], were not significant in our sample. This may indicate sampling or population differences but may also reflect measurement issues. While the combining of grandiosity and vulnerability in the PNI is intended to reflect clinical presentations of narcissistic dysfunction, features such as vanity and haughtiness may be underrepresented. Since the PNI does not clearly distinguish between grandiosity and vulnerability from one another [33], future studies examining these specific components should employ multiple measures of narcissism. It is also possible that the brevity of the brief IPIP agreeableness dimension failed to capture the breadth of narcissism-related antagonism, pointing to a need for broader five-factor measurement in future research. Other limitations of the present study include the use of an online community sample with lower levels of narcissistic dysfunction than in clinical settings, and the exclusive use of self-report assessment at a single time-point, including asking for retrospective perceptions of adversity in childhood. Indeed, the cross-sectional design carries the risk of shared method variance, and precludes inferences regarding directionality and causation.

Despite the limitations of this preliminary study, findings highlight directions for further research and indicate important personality domains for potential clinical exploration among individuals with pathological narcissism, who are often misunderstood and under-treated.