Level of personality functioning and maladaptive personality traits in relation to depression and anxiety symptoms in middle and older adults

Abstract

A growing body of research suggests that personality pathology underpins various mental disorders and serves as a risk factor for developing emotional disorders such as depression and anxiety. This study examined how levels of personality functioning and pathological traits predict the severity of these symptoms in a community sample of middle and older adults (N = 530). The Generalized Anxiety Disorder Questionnaire, Patient Health Questionnaire, Self and Interpersonal Functioning Scale, and Personality Inventory for ICD-11 were used. Regression analyses revealed that impaired identity and empathy significantly predicted anxiety symptoms, while only identity emerged as a predictor of depression. Self-direction and intimacy did not significantly predict either disorder. Negative Affectivity and Detachment were key predictors of anxiety, while depression was primarily related to Negative Affectivity. Variance partitioning analysis demonstrated that personality functioning and maladaptive traits independently and jointly contribute to symptom severity, with overlapping variance highlighting their interconnected roles in emotional pathology. These findings underscore the importance of assessing both personality functioning and traits when addressing emotional disorders, particularly during middle and late adulthood, a stage characterized by unique psychological and social challenges. Integrating personality assessments into routine mental health care can enhance diagnostic accuracy, improve treatment planning, and optimize outcomes for individuals with depression and anxiety.

Introduction

A growing body of research indicates that personality pathology may underlie various other mental disorders and serve as a risk factor for the onset of emotional disorders (ED), such as depression or anxiety disorders^1–5. This perspective is based on the notion that personality pathology can create a vulnerability to emotional disorders or other forms of psychopathology, what aligns with what is often referred to as the predisposition model. It has a long-standing reputation in research but is just one of several approaches (i.e., common cause, continuum/spectrum, precursor, predisposition, pathoplasticity, concomitants, and consequences/scar models) that offer frameworks for examining the relationship between personality and psychopathology (see for a review⁶). According to Klein et al.⁷, these models can be grouped into three categories (1) models that view personality and EDs as sharing common causal influences without assuming a direct causal relationship (common cause, continuum/spectrum, and precursor); (2) models suggesting that personality exerts a causal effect on the onset or maintenance of EDs (predisposition and pathoplasticity); and (3) models proposing that EDs have a causal influence on personality (concomitants and consequences). One of the key questions in this context is which of these models best captures the nature of the associations between personality and psychopathology. For instance, Tackett’s⁸ review within the framework of developmental psychopathology indicated that preliminary evidence supports the predisposition model alongside the spectrum approach. While the predisposition model remains widely adopted, the spectrum approach has gained significant traction in recent years, particularly in personality disorder research. However, it has also faced substantial criticism^9,10. Nevertheless, it is likely that all models contribute to a comprehensive understanding of personality–psychopathology associations, potentially varying in relevance depending on the specific type of psychopathology examined⁸. In our study, this perspective translates into an effort to determine the specificity of particular ED symptoms, depression and generalized anxiety. It is also worth noting that research on the personality–psychopathology relationship has traditionally studied healthy or pathological traits¹¹. However, an alternative conceptualization—viewing personality as an organization of self and interpersonal functioning—has received relatively less attention. This study addresses this gap by examining personality vulnerability not only in terms of pathological traits but also in relation to personality functioning (identity, self-direction, empathy, and intimacy). A more comprehensive conceptualization of personality may lead to different findings than those typically observed in trait-based research.

This has become even more apparent since the introduction of a dimensional model of personality disorder (PD) diagnosis in a completely redefined definition of PD in ICD-11 ¹² and earlier in the Alternative Model of Personality Disorders (AMPD) in Section III of DSM-5¹³. A key feature of AMPD is that the diagnosis of personality disorder requires the assessment of both personality functioning and pathological traits. The parallel dimensional model of personality disorder diagnosis in ICD-11 is, on the one hand, compatible with AMPD, as it incorporates analogous core features of PD (self and interpersonal impairments) and is supported by empirical evidence demonstrating their similarity (see, e.g^14,15). On the other hand, there are notable differences between the two models of personality disorder assessment. In ICD-11, only the first step—assessing the severity of personality pathology—is mandatory, whereas the evaluation of pathological traits and the borderline pattern specifier is optional. Additionally, the composition of pathological traits differs between the systems: DSM-5 AMPD utilizes five broad trait domains (Negative Affectivity, Detachment, Antagonism, Disinhibition, Psychoticism), while ICD-11 includes an alternative configuration (Negative Affectivity, Detachment, Dissociality, Disinhibition, Anankastia), with the primary distinction being the inclusion of anankastia in ICD-11. Notably, research suggests that the ICD-11 model, which accounts for Anankastia, provides a more comprehensive representation of personality pathology than the DSM-5 model¹⁶, leading to recommendations that future iterations of AMPD incorporate Anankastia as well¹⁷. Another key distinction lies in how personality functioning criteria are articulated. ICD-11 places greater emphasis on behavioral manifestations (e.g., self-harm or externalized psychosocial dysfunction¹⁸), whereas DSM-5 AMPD focuses more on intrapsychic and regulatory processes, such as identity and intimacy. While the ICD-11 approach to diagnosis is arguably less psychologically nuanced, making it more accessible to a broader range of practitioners, its status as the official classification system endorsed by the World Health Organization means that it will be fully implemented in Poland’s healthcare system by the end of 2026. Given this impending transition, advancing research utilizing elements of the ICD-11 model represents a particularly relevant and pressing practical challenge.

Research supports the importance of assessing personality functioning beyond the scope of personality disorders^2,3,19, as treatment failures in addressing symptom disorders may stem from comorbid personality disorders or subclinical manifestations of personality pathology. Indeed, personality pathology has been shown to complicate standard treatment outcomes in treating anxiety or depression^3,20. Hence, it is suggested that the assessment of the level of personality functioning should be considered mandatory for every psychiatric patient as part of routine screening procedures^2,19,21. However, there is still limited understanding of the nature of personality functioning impairment in other mental health conditions, although research on these relationships is still growing. Emotional disorders are among the most prevalent comorbid disorders in populations of patients with personality disorders²². This association becomes even more apparent in the context of adverse childhood experiences: recent studies have identified a mediating effect of personality functioning in the relationship between childhood trauma and both depression and anxiety symptoms^5,23,24. This body of research demonstrates the clinical relevance of assessing level of personality functioning in ED, revealing that some patients exhibit severe underlying personality dysfunction, while others exhibit only mild or no impairment (see review³). In general, individuals with emotional disorders show more impaired personality functioning compared to healthy controls, but those with personality disorders tend to exhibit even higher levels of impairment than those with anxiety and mood disorders alone². Furthermore, no significant differences have been observed between individuals with anxiety and mood disorders in terms of their level of personality functioning, nor between various anxiety disorders. However, some studies have not found a significant relationship between personality functioning and anxiety symptoms^25,26. This highlights the need for further research to clarify the importance of personality functioning in relation to anxiety and mood disorders, especially in understudied less convenient samples and using a new dimensional approach model to the assessment of PD.

Together with the level of personality functioning, the assessment of maladaptive personality traits is crucial to understanding the relationship between personality pathology and emotional disorders. A recent meta-analysis by Gioletti and Bornstein²⁷ used a meta-analytic approach to confirm that Personality Inventory for DSM-5 (PID-5) trait scores predict symptom disorders, including anxiety and depression, showing that all five personality traits were related to both ED, with particularly high effect sizes for Negative Affectivity and Detachment. Bach et al.²⁸ found that maladaptive personality traits—Negative Affectivity, Detachment, and Psychoticism—mediate the relationship between childhood trauma and adult internalizing symptoms (both anxiety and depression). Additionally, Hong et al.²⁹ demonstrated that DSM-5 personality traits, Negative Affectivity, and Detachment are related to cognitive risks for depression, anxiety, and obsessive-compulsive symptoms, highlighting a core transdiagnostic factor that links maladaptive traits to various cognitive vulnerabilities. Further research by Komasi et al.³⁰ and Vittengl²⁰ has shown that depressive symptoms are strongly associated with internalizing trait dimensions, particularly increased Negative Affectivity or high Neuroticism, and increased Detachment. There is also a lesser but notable correlation with other maladaptive traits, such as elevated Disinhibition/low Conscientiousness, as well as greater levels of Psychoticism and Antagonism²⁰. In a study of 640 outpatients with mood and anxiety disorders using the NEO Five-Factor Inventory, Cuijpers et al.³¹ identified significant differences based on the number of comorbid disorders rather than specific types, with notable variations in neuroticism and agreeableness among patients with increasing comorbidity. While the NEO Five-Factor Model assesses adaptive personality traits, these findings remain relevant as research indicates that maladaptive traits often serve as exaggerated or dysfunctional counterparts to adaptive traits, influencing psychological well-being and psychopathology³². This body of research highlights the importance of maladaptive personality traits in understanding the complexity of emotional disorders, although it remains unclear whether these relationships are specific to individual disorders and whether they persist when ICD-11 trait model is applied. Further research is needed to explore whether anxiety and depression symptoms are differently related to maladaptive personality traits and whether these traits contribute additional variance beyond the level of personality functioning as predictors of emotional disorders.

Thanks to dimensional models, we can go beyond the mere comorbidity of descriptive diagnoses and delve deeper into the underlying mechanisms and interrelations among different psychopathological constructs, thereby enhancing our understanding of the complexity of mental health disorders. This is particularly important to study age groups beyond young adults in this approach, as despite the emphasis of developmental psychopathology on understanding the nature and course of PD throughout life^33,34, most research on PD has mainly focused on early adulthood (ages 18–30). Consequently, there is still a lack of clarity regarding how PD manifests during middle and late adulthood. Since categorical diagnoses were developed primarily based on a prototypical younger adult, a dimensional approach may be more relevant for studying middle-aged and older individuals³⁵. Research also suggests that assessments of personality functioning and personality traits exhibit little measurement bias across age groups, making these assessments more robust across the lifespan³⁶. Individuals beyond their 30s face specific challenges, such as changes in career, interpersonal relationships, and physical health, which may influence the severity and flavor of psychiatric symptoms³⁵. They also may experience existential crises related to assessing life achievements and future plans³⁷. Such factors may modify the way personality pathology relate to emotional disorders in this life stages. It is also important to note that there is a limited number of studies exploring new dimensional model of personality disorders in this age groups, highlighting the need for further research in this area.

In conclusion, assessing personality pathology in the context of anxiety and depressive symptoms is emphasized by both clinicians and empirical research, which demonstrates a significantly increased risk of developing depression^38,39 as well as anxiety disorders⁴⁰ among individuals with personality disorders. While there is a growing body of research exploring the relationship between emotional disorders and PD, only a few focused on the new dimensional models of PD, with one study that examined the interaction between level of personality functioning and maladaptive personality traits in explaining emotional symptoms²⁰ and no studies within the ICD-11 framework. Additionally, most existing studies focus on young adults or adult samples aged 18–60, with the mean age often falling within the younger adult range. There is a critical need to better understand changes in patterns of personality pathology and mental health over the entire lifespan. Expanding research beyond samples of convenience, primarily young adults, can significantly enhance our understanding of the role of personality functioning and maladaptive personality traits in explaining ED, both depression and anxiety symptoms. Gaining insight into how these disorders manifest and change as individuals age is key to developing effective treatments and improving the quality of life for people beyond young adults affected by these conditions.

Aims of the study

This study aimed to investigate the relationships between the level of personality functioning, maladaptive personality traits, and symptoms of depression and anxiety in a community sample of middle and older adults. We formulated two hypotheses.

We hypothesized that dimensions of personality functioning (impairments in identity, self-direction, empathy, and intimacy) would be positively associated with the severity of both depression and generalized anxiety symptoms.

We also expected that pathological traits would be related to these emotional disorders, with stronger associations for internalizing traits (Negative Affectivity, Detachment) and weaker associations for externalizing traits (Disinhibition, Dissociality).

Furthermore, considering the diagnostic framework proposed in ICD-11 and DSM-5 AMPD, which recommends a two-step approach for comprehensive personality disorder conceptualization and assessment—evaluating both the severity of psychopathology and pathological traits—we posed the following research question: To what extent do the level of personality functioning and pathological traits uniquely and jointly predict the level of symptoms of depression and anxiety?

We hypothesized that both the level of personality functioning and pathological traits would uniquely and jointly predict the level of symptoms of depression and anxiety. Specifically, we expected that while dimensions of personality functioning would uniquely account for variance in these symptoms, pathological traits would provide additional unique variance and some overlap with personality functioning in predicting these symptoms.

Methods

Transparency and openness

We report how we determined our study sample and how we handled missing data, all manipulations, and all measures in the study. Adhering to the Journal Article Reporting Standards⁴¹, all the data and research materials are available online. This study was not preregistered.

Participants and procedure

The study sample consisted of 530 participants, Polish residents with a nearly equal distribution of females (51.5%) and males (48.1%), and a small representation of other gender identities (0.4%). The mean age of participants was 47.4 years (SD = 12.4), ranging from 30 to 97 years. 75% of the sample comprised middle-aged adults (30–54 years old), while the remaining 25% were late adults (55 years and older). The majority of participants had higher education, with 45.7% holding a BA degree and 14.5% holding an MA degree, while 39.8% had completed secondary education. Regarding marital status, 61.7% were married, 15.8% were in informal relationships, 10.9% were single, 7.7% were divorced or separated, and 3.8% were widowed. Regarding occupation, 64.5% were employed full-time, 5.7% part-time, and 16.0% were retired. Additionally, 34.7% of the participants had received psychotherapy or some form of psychological help at some point in their lives, while 7% had undergone outpatient or hospital treatment for mental health difficulties (see Table 1 for demographic characteristics of the sample).

Table 1.

Demographic characteristics of the sample (N = 530).

Variable	N	%
Female	273	51.5
Male	255	48.1
Other	2	0.4
Education
Secondary	211	39.8
Higher (BA)	77	14.5
Higher (MA)	242	45.7
Marital status
Single	58	10.9
Informal relationship	84	15.8
Marriage	327	61.7
Divorced or separated	41	7.7
Widowed	20	3.8
Occupation
During training/education	2	0.4
Unemployed	22	4.2
Homemaker/Stay-at-home parent	34	6.4
Full-time employment	342	64.5
Part-time employment	30	5.7
Retirement	85	16.0
Disability pension	15	2.8
Received psychotherapy or psychological help (lifetime. yes)	184	34.7
Outpatient or hospital treatment for mental health (lifetime. yes)	36	6.8

Participants were recruited via the Ariadna research panel, a professionally managed online platform. The panel’s systematic approach to data collection resulted in a dataset with no missing values, ensuring the robustness and completeness of the analyses. It is important to note that participants in this type of panel receive points for completing questionnaires, which can be exchanged for rewards. This incentive structure may have influenced their motivation to participate in the research reported here. However, the panel also includes built-in checks for attention, and the completion time is monitored. Additionally, prior to our analyses, we conducted a thorough data review and excluded participants who exhibited response patterns indicative of ‘answering styles’ (e.g., consistently selecting only the middle option). Thus, a total of 67 observations were removed from the dataset. All participants gave written informed consent for their participation in this research. All methods were performed in accordance with relevant guidelines and regulations, including ethical principles outlined in the Declaration of Helsinki. The study protocol was approved by the Human Research Ethics Committee at Faculty of Psychology and Cognitive Science at Adam Mickiewicz University in Poznan, Poland decision no. 5/11/2024.

Measures

The Generalised Anxiety Disorder Questionnaire (GAD-7⁴²) is a screening tool for assessing the risk of generalized anxiety syndrome. It consists of 7 items addressing various GAD syndrome symptoms in the past two weeks. Participant addresses the questions on a four-point scale, ranging from 0="not bothered at all” to 3="almost every day”. The total score is obtained by summing the responses to all seven items, ranging from 0 to 21. A higher score indicates greater severity of symptoms of GAD. We used Polish, validated⁴³ freely available version of the questionnaire prepared by the MAPI Research Institute (www.phqscreeners.com). The scale has excellent internal consistency (α = 0.94).

The Patient Health Questionnaire (PHQ-9⁴⁴) is a self-report tool used for the screening assessment of the risk of depressive disorders. It consists of 9 questions and an additional question about the severity of symptoms. The questions inquire about the frequency of depressive symptoms (according to diagnostic criteria based on DSM-IV) experienced in the past two weeks. Participants respond on a four-point scale, where 0 indicates “not at all,” 1 - “several days,” 2 - “more than half the days,” and 3 - “nearly every day.” The total score is obtained by summing the responses to all seven items, ranging from 0 to 27. A higher score indicates a greater severity of depression symptoms. The scale has excellent internal consistency (α = 0.91). We used the Polish version of PHQ-9, validated in a sample of middle-aged and older adults⁴⁵.

The Self and Interpersonal Functioning Scale (SIFS) by Gamache et al.⁴⁶ in the Polish adaptation⁴⁷, was utilized to assess the level of personality pathology. The scale comprises 24 items, rated on a 5-point Likert scale ranging from 0 (does not describe me at all) to 4 (describes me completely accurately). The total score reflects the overall level of personality dysfunction. Domain scores are computed separately for: identity (α = 0.85), self-direction (α = 0.61), empathy (α = 0.76), and intimacy (α = 0.75). Higher scores indicate greater impairment in personality functioning. All subscales, except self-direction (which falls within an acceptable range), demonstrated good internal consistency. The SIFS is a valid and clinically applicable tool that aligns with the ICD-11 framework⁴⁸.

The personality inventory for ICD-11 (PiCD⁴⁹) is a 60-item self-report instrument assessing the five ICD-11 personality trait domains: Negative Affectivity, Detachment, Dissociality, Disinhibition, and Anankastia. Participants rate their agreement to items on a scale from 1 (strongly disagree) to 5 (strongly agree). The total score for each trait domain is calculated as the mean of the 12 items corresponding to that domain. Higher scores indicate greater presence of maladaptive personality traits associated with personality disorders. The PiCD has been validated in a Polish sample, showing satisfactory internal consistency, factorial validity, and convergent-discriminant validity⁵⁰. In the current study, Cronbach’s alpha reliability was as follows: Negative Affectivity - α = 0.91, Detachment - α = 0.89, Dissociality - α = 0.87, Disinhibition - α = 0.89, and Anankastia - α = 0.75.

Statistical analyses

Pearson’s correlation coefficients were calculated to assess the strength and direction of associations between variables. To investigate to what extent dimensions of personality functioning and pathological traits uniquely and jointly predict dependent variables (symptoms of depression and generalized anxiety), we conducted hierarchical multiple regression with a stepwise procedure for each emotional disorder separately. Gender and age were included as control variables to assess whether dimensions of personality functioning and pathological traits remained significant predictors of depression and anxiety. Incremental validity was evaluated by examining the added value of each group of variables when introduced in separate steps. Collinearity was assessed using the variance inflation factor (VIF), which in our study did not exceed 3.37, indicating no issues with multicollinearity. Additionally, autocorrelation (Durbin-Watson), Cook’s distance, and normality of distribution tests were performed. To delineate the unique and shared variance explained by the dimensions of personality functioning and pathological traits, we conducted a variance partitioning analysis (VPA).

Results

Gender and age differences in personality and symptom severity

Mean scores between the groups of women and men were compared for the studied variables (after filtering out individuals who identified their gender as ‘other’) using Welch’s t-test. The results show significant between-group differences across all personality-related variables (except for Disinhibition, Anankastia, and the overall personality functioning score), as well as the severity of generalized anxiety. Women were found to have higher impairments in Identity (t (521) = 2.30, p = .02, d = 0.20), Self-direction (t (522) = 2.28, p = .02, d = 0.19), and generalized anxiety (t (526) = 2.25, p = .025, d = 0.20) and lower impairments in Empathy (t (509) = -3.53, p < .001, d = 0.31), Intimacy (t (525) = -3.13, p = .002, d = 0.27), Detachment (t (525) = -2.91, p = .004, d = 0.25), and Dissociality (t (510) = -6.70, p < .001, d = 0.58). We also examined the relationships between age and personality variables, as well as the severity of symptoms in our study. Age was negatively correlated with impairments in personality functioning and with pathological traits (except for Anankastia), with correlations ranging from r = − .27, p < .001 for identity impairments to r = − .10, p < .05 for Detachment. Overall, this indicates that older individuals tend to exhibit fewer personality impairments and symptoms. In subsequent analyses, gender and age were included as control variables.

Correlations between personality characteristics and emotional disorder symptoms

Table 2 presents correlations between generalized anxiety and depression symptoms and personality characteristics. The results indicate significant positive correlations between anxiety and depression (r = .81, p < .001), suggesting a strong relationship between the symptoms of these emotional disorders. Additionally, the level of personality functioning (overall SIFS) (r = .64, p < .001), as well as subscales: identity (r = .65, p < .001), and self-direction (r = .50, p < .001) showed moderate correlations with anxiety, with similar patterns observed with depression. Empathy (r = .50, p < .001) and intimacy (r = .40, p < .001) were also moderately correlated with anxiety, while personality traits like Negative Affectivity (r = .70, p < .001) and Disinhibition (r = .46, p < .001) had moderate to strong correlations with anxiety and depression. In terms of age, negative correlations were observed with anxiety (r = − .22, p < .001) and depression (r = − .14, p < .001), indicating that older individuals reported lower levels of anxiety and depression.

Table 2.

Descriptive statistics and correlations between personality characteristics and symptoms of emotional disorders (N = 530).

	M	SD	Minimum–maximum	Skew. (a)	Kurt.(b)	Shapiro-Wilk W	Age	GAD7	PHQ9
Generalized anxiety	6.42	5.46	0–21	0.75	− 0.16	0.92***	− 0.22***	–	–
Deppresion	7.31	6.18	0–27	0.96	0.41	0.91***	− 0.14***	0.81***	–
Personality functioning	1.28	0.58	0-3.17	0.44	-0.41	0.98***	− 0.27***	0.64***	0.64***
Identity	1.39	0.74	0-3.43	0.47	-0.62	0.96***	− 0.27***	0.65***	0.66***
Self direction	1.43	0.67	0-3.80	0.19	-0.34	0.98***	− 0.19***	0.50***	0.51***
Empathy	1.10	0.72	0–4	0.63	-0.14	0.95***	− 0.15***	0.50***	0.45***
Intimacy	1.19	0.75	0-3.83	0.48	-0.28	0.97***	− 0.24**	0.40***	0.40***
Negative affectivity	34.00	9.31	12–16	-0.12	-0.32	0.99**	− 0.14**	0.70***	0.63***
Disinhibition	25.50	8.35	12–57	0.62	0.08	0.97***	− 0.13**	0.46***	0.48***
Detachment	29.70	8.87	12–53	0.15	-0.64	0.99***	− 0.10*	0.33***	0.39***
Dissociality	25.00	7.81	12–55	0.65	0.21	0.97***	− 0.15***	0.34***	0.32***
Anankastia	40.50	5.73	17–56	-0.35	0.97	0.99***	0.08	0.06	0.04

*** p < .001; ** p < .01; one-tailed; (a) Std. error skewness; 0.106; (b) Std. error kurtosis 0.212.

Incremental validity of personality functioning and maladaptive personality traits in predicting emotional disorders

Hierarchical regression analysis for generalized anxiety controlling for gender and age

In predicting the severity of generalized anxiety symptoms, the first step of the hierarchical regression analysis included the variables representing the four dimensions of personality functioning (see Table 3). The resulting model was statistically significant and explained 46% of the variance in symptoms (Adj. R² = 0.46; F(4, 523) = 114.7, p < .001). Significant predictors of generalized anxiety severity were impairments in Identity (β = 0.50; p < .001), Self-direction (β = 0.10; p = .019), and Empathy (β = 0.19; p < .001). In the second step, a set of pathological traits was introduced. The predictive power of the model increased significantly by 9% (ΔR² = 0.09; F(5, 518) = 20.72, p < .001; total 54%). Both Identity (β = 0.27; p < .001) and Empathy impairments (β = 0.15; p < .001) remained significant predictors, but the significance of Self-direction diminished. Significant predictors also included Negative Affectivity (β = 0.45; p < .001) and Detachment (β = − 0.11; p = .014). In the third step, age and gender were added as control variables, which slightly increased the percentage of explained variance by 0.5%, though this difference between models was not statistically significant. In this model, age emerged as a statistically significant predictor (β = − 0.07; p = .023).

Table 3.

Hierarchical linear regression testing effects of personality characteristics on generalized anxiety.

Step	Δ Adj. R2	F-change	β	t-value	95% CL
					LL	UL
Step 1	0.463	114.7***
Identity			0.25	5.49***	1.20	2.53
Self direction			0.02	0.36	− 0.58	0.84
Empathy			0.16	3.49***	0.52	1.86
Intimacy			0.01	0.30	− 0.54	0.73
Step 2	0.088	20.72***
Negative affectivity			0.45	8.96***	0.21	0.32
Disinhibition			− 0.01	− 0.21	− 0.08	0.06
Detachment			− 0.10	− 2.33*	− 0.11	− 0.01
Dissociality			0.03	0.57	− 0.04	0.08
Anankastia			0.02	0.63	− 0.05	0.09
Step 3	0.005	2.98
Age			− 0.07	− 2.29*	− 0.06	0.00
Sex			− 0.04	− 0.60	− 0.91	0.49

The coefficients shown are those in the final model that accounted for all other predictors. We report 95% CIs for beta coefficients. The gender was coded − 1 for women and 1 for men.

*p < .05. **p < .01. ***p < .001.

Hierarchical regression analysis for depression controlling for gender and age

In predicting the severity of depression symptoms, the first step of the hierarchical regression analysis included the variables representing the four dimensions of personality functioning (see Table 4). The resulting model was statistically significant and explained 47% of the variance in symptoms (Adj. R² = 0.47; F(4, 523) = 115.5, p < .001). Significant predictors of depressive symptoms severity were impairments in Identity (β = 0.38; p < .001). In the second step, a set of pathological traits was introduced. The predictive power of the model increased significantly by 4% (ΔR² = 0.04; F(5, 518) = 7.22, p < .001; total 50%). Identity impairments (β = 0.38; p < .001) remained a significant predictor. The next significant predictor was the Negative Affectivity (β = 0.21; p < .001). In the third step, age and gender were added as control variables, which increased only marginally the percentage of explained variance by less than 0.1%, and this difference between models was not statistically significant.

Table 4.

Hierarchical linear regression testing effects of personality characteristics on depression.

Step	Δ Adj. R2	F-change	β	t-value	95% CL
					LL	UL
Step 1	0.465	115.5***
Identity			0.38	7.66***	2.34	3.95
Self direction			0.08	1.72	-0.11	1.60
Empathy			0.05	1.14	-0.34	1.28
Intimacy			− 0.02	-0.33	-0.89	0.64
Step 2	0.035	7.22***
Negative affectivity			0.21	3.96***	0.07	0.21
Disinhibition			0.08	1.42	-0.02	0.14
Detachment			0.04	0.90	-0.03	0.09
Dissociality			− 0.01	-0.13	-0.09	0.07
Anankastia			0.04	1.05	-0.04	0.13
Step 3	0.001	0.67
Age			− 0.03	-0.91	-0.05	0.02
Sex			− 0.04	-0.60	-1.10	0.58

The coefficients shown are those in the final model that accounted for all other predictors. We report 95% CIs for beta coefficients. The gender was coded − 1 for women and 1 for men.

*p < .05. **p < .01. ***p < .001.

Incremental validity: variance partitioning analysis

To examine to what extent dimensions of the level of personality functioning versus pathological traits uniquely predict depression (PHQ-9) and anxiety (GAD-7), we performed a variance partitioning analysis (VPA). The results indicate that both dimensions of personality functioning (32%) and the set of pathological traits (35.1%) make a significant unique contribution to explaining the variance in the severity of generalized anxiety symptoms, with pathological traits accounting for a slightly larger portion of the variance than the dimensions of personality functioning. Additionally, there is a substantial portion of variance (14.7%) that is shared between both predictors, suggesting that the dimensions of personality functioning and pathological traits have some overlap in predicting the severity of generalized anxiety symptoms.

In the case of explaining the variance in the severity of depressive symptoms, the dimensions of personality functioning account for a larger portion of unique variance (34.8%) compared to the set of pathological traits (29.7%). However, similar to generalized anxiety, there is a shared variance (12.1%) between the dimensions of personality functioning and pathological traits, indicating that these two variables partially overlap in explaining depressive symptoms.

Discussion

The aim of this study was to examine the relationships between level of personality functioning, pathological traits, and symptoms of depression and anxiety in a community sample of middle and older adults. We hypothesized that impairments in key dimensions of personality functioning (identity, self-direction, empathy, and intimacy) would be positively associated with the severity of both depressive and generalized anxiety symptoms. Additionally, we anticipated that pathological personality traits would be linked to these emotional disorders, with stronger associations expected for internalizing traits (such as Negative Affectivity and Detachment) and weaker associations for externalizing traits (such as Disinhibition and Dissociality). Moreover, we sought to investigate the extent to which levels of personality functioning and pathological traits uniquely and jointly predict symptoms of depression and anxiety. This was the first study to investigate these relationships within the ICD-11 framework for personality disorder diagnosis.

Regarding the dimensions of personality functioning, although the correlations between all four scales of personality functioning and both ED were consistently moderate to strong, regression analyses revealed more nuanced associations: impaired identity and empathy emerged as significant predictors of anxiety symptoms, while for depression, only identity was a significant predictor. Interestingly, self-direction and intimacy did not emerge as significant predictors in the regression models for either depression or anxiety symptoms. This finding suggests that, while these dimensions are central to overall personality functioning, they may play a less direct role in the specific symptomatology of emotional disorders. In terms of pathological traits, our hypothesis was fully supported for anxiety symptoms, with Negative Affectivity and Detachment as significant predictors, and partially confirmed for symptoms of depression, where Negative Affectivity was the significant predictor. Thus, certain aspects of personality pathology are important for both emotional disorders, jointly explaining 46% of the variance in anxiety and 49% in depression. These findings demonstrate that both impairments in personality functioning and pathological personality traits, which are important components of dimensional models of PD diagnosis in ICD-11 and DSM-5 AMPD, are indicative not only of personality pathology but also of symptom disorders²⁷, which is in line with other studies of dimensional and categorical diagnosis of PD for both anxiety^2,4,40, and depression^51,52. Our study extends these findings by showing that approximately 50% of the symptoms of emotional disorders can be related to the level of personality functioning and maladaptive personality traits, measured within the ICD-11 framework. Personality pathology, being more stable over time than emotional disorders, may serve as a foundational substrate upon which ED symptoms manifest and fluctuate^3,4. Even in the absence of a full-blown personality disorder, patients with symptoms of emotional disorders can exhibit significant impairments in personality functioning and pathological traits. Consequently, treating such patients should account for the presence of personality-related issues, as addressing these problems is crucial for comprehensive care.

Our study revealed that both identity impairments and high Negative Affectivity were associated with symptoms of emotional disorders. Negative Affectivity has also been linked to ED in other studies^27,29,30. This domain likely contributes to the emotional difficulties observed in ED, as it closely mirrors symptoms of anxiety and depression, involving frequent and intense experiences of negative emotions, such as anxiety, depression, guilt, and anger. It also encompasses emotional instability, heightened nervousness, worries, and fear, alongside separation insecurity and fear of rejection or separation from significant others. Identity refers to the experience of oneself as a unique individual, characterized by clear self-other differentiation, stable and adequate self-esteem, and the ability to regulate a wide range of emotions¹³. Identity integration is often regarded as a fundamental aspect of a healthy personality⁵³. The link between identity and emotional disorders can be understood through developmental psychoanalytic theories, which highlight the role of early identity formation in shaping various forms of psychopathology⁵⁴. Nevertheless, identity development is a lifelong process that continues to evolve throughout middle and late adulthood. While core aspects of personality remain relatively stable, environmental and situational factors—such as changing family dynamics, career shifts or retirement, and loss of significant relationships—may impact self-perception and emotional adaptation⁵⁵. Theoretical perspectives, such as Erikson’s concept of generativity versus stagnation and later integrity versus despair, suggest that maintaining a coherent identity is essential for psychological well-being at different points in adulthood. Difficulties in adapting to these transitions may contribute to identity disturbances, emotional distress, and increased vulnerability to psychopathology⁵⁶. By incorporating this lifespan perspective, our findings suggest that identity impairments in middle and late adulthood are contextually influenced vulnerabilities that shape the severity of emotional disorders across adulthood. Further longitudinal studies should explore how identity-related difficulties in midlife and later adulthood may contribute to the development of emotional problems, shaping mental health outcomes across the lifespan.

However, in our study, impaired empathy and low Detachment also emerged as significant predictors, but only for anxiety symptoms. This suggests that anxiety has a strong interpersonal aspect, which is less pronounced in depression. Empathy is part of interpersonal dimension of personality functioning, and Detachment is a trait that regulates interpersonal distance by withdrawal from relationships due to perceived interpersonal threats¹³. These results are consistent with other studies, which indicate that the relationship for self-pathology is stronger in depression, whereas interpersonal impairments are equally significant in anxiety^20,57. Impaired empathy may significantly contribute to anxiety symptoms, as maladaptive mindreading and reduced understanding of others’ experiences can reinforce anxiety, leading to avoidance of social interactions and perceiving them as threatening. This, in turn, can exacerbate interpersonal difficulties and further deepen empathy deficits. These findings align with other studies showing deficits in mentalizing⁵⁸ and different aspects of empathy⁵⁹ among individuals with anxiety disorders. Contrary to previous studies, regression analysis showed that a lower level of Detachment was associated with higher levels of anxiety, even though the initial correlations between these two variables were positive. This suggests that individuals who avoid relational experiences and intimacy, prefer solitude, and are highly suspicious of others may actually be less prone to experiencing anxiety symptoms. It is possible that high Detachment may reduce the experience or reporting of generalized anxiety symptoms by potentially decreasing sensitivity to social stimuli. Since many symptoms of generalized anxiety appear to involve worrying about relationships and other people, this area might be less prominent or less consciously perceived by individuals with high Detachment. This aligns with the broader interpersonal aspect of anxiety (that is less evident in depression), where deficits in understanding and engaging with others can contribute to symptom severity.

Our final aim was to examine the extent to which personality functioning and pathological traits uniquely and jointly predict symptoms of depression and anxiety. This investigation builds on the extensive literature addressing the incremental validity of both indices of personality pathology in predicting critical clinical outcomes^60–62. The results indicate that all personality functioning dimensions and pathological traits significantly contribute to explaining the variance in generalized anxiety severity, with pathological traits accounting for a slightly larger portion (35.1% vs. 32%). Additionally, 14.7% of the variance is shared between these predictors, indicating some overlap. For depressive symptoms, personality functioning dimensions explain a larger portion of the unique variance (34.8%) compared to pathological traits (29.7%), with 12.1% of the variance being shared. Our findings support the notion that personality functioning, and maladaptive traits represent distinct aspects of personality pathology, each uniquely contributing to clinical outcomes such as emotional disorders. Notably, symptoms of depression and anxiety differ in their associations with these predictors: generalized anxiety disorder is more strongly predicted by pathological traits, while depression is more closely related to levels of personality functioning, albeit with relatively small specificity. The shared variance of 12–15% underscores the interconnectedness of these constructs, indicating that while they are distinct, there is some overlap in their contribution to emotional disorder symptoms. This overlap between level of personality functioning and maladaptive traits may also reflect potential redundancy between those two criteria. Indeed, empirical studies have demonstrated strong arguments for the overlap between Criterion A and B of AMPD, as evidenced by empirically oriented researchers⁶³, while this notion has been contested by clinically oriented scholars^9,10,64. As far as we know, our study is the first to examine incremental variance in ED using the ICD-11 model, while the majority of studies examining the relationship between personality functioning (pathology/severity) and personality traits are based on the DSM-5 AMPD framework (e.g., Martí Valls et al., 2023; Nysaeter et al., 2023). This debate is particularly relevant given the predominant reliance on self-report measures⁶⁵, which contributes, at least in part, to a challenge posed by the hard-to-mitigate common method variance⁶⁶. Consequently, future research should not only incorporate data collection methods beyond self-report in assessing both personality functioning and pathological traits but should also—following the principles of multimethod assessment and clinical utility approaches—engage in a critical reflection on the clinical goals that can or cannot be effectively addressed using those two elements of dimensional model of PD (e.g., nosological description vs. case conceptualization).

Our finding of 12% shared variance between personality functioning and maladaptive traits differs significantly from Vittengl et al.²⁰, who reported that more than 40% of the predictive power of personality pathology for depression was shared, suggesting that the constructs of personality functioning and traits largely overlap in explaining depression. This discrepancy may stem from several methodological and conceptual differences between these studies. First, Vittengl et al.’s sample included a large proportion of clinical participants with various psychiatric diagnoses, including clinical depression, whereas our sample was community-based. Thus, our findings suggest that in a less severe and more diverse population, personality dysfunction and maladaptive traits contribute to depression in a more independent manner. In contrast, in clinical samples, the greater shared variance may reflect a stronger general psychopathology factor (g-factor), potentially amplified by symptom severity and comorbidity⁶⁷. Second, we used ICD-11-based measure for traits, whereas Vittengl et al. employed DSM-5 AMPD-based assessments. Those models differ in their emphasis on the distinction between personality dysfunction and traits: the AMPD model is designed to emphasize their integration, whereas the ICD-11 model conceptualizes maladaptive traits and personality functioning as more distinct and independent. Third, Vittengl et al. used a structured clinical interview to diagnose depression, while our study relied on self-report measures. Structured interviews typically yield more reliable, clinically validated diagnoses, potentially capturing stronger associations between personality pathology and depression. However, self-report measures may be more sensitive to subjective experiences of distress and subthreshold symptomatology, which could lead to greater differentiation between personality functioning and trait-based predictors in our study.

Our study provides some additional evidence suggesting that personality pathology is strongly related to emotional disorders even in midlife and late adulthood, as the mean age of participants in our sample was 47 years, with a minimum age of 30. The period after 30 years of age may represent a particularly critical phase for the manifestation of psychopathological symptoms, as individuals may encounter escalating social, physical, and cognitive challenges³⁵. Those with pre-existing maladaptive personality patterns may experience greater difficulties than their peers, as evolving social contexts and increasing demands exacerbate their struggles to cope. This period may mark the first time that such an individual’s ability to navigate life’s challenges diminishes, potentially leading to significant impairments in daily functioning and a heightened risk for emotional disorders. It is possible that ED might emerge as a consequence of underlying personality pathology during this life stage⁶⁰. Early identification of PD traits could be crucial in predicting and potentially preventing the development of ED in later life⁶⁸. However, in our community sample, age was weakly but significantly and negatively correlated with all psychopathological measures. This could be indicative of an age-related bias in the way these symptoms are formulated, potentially failing to capture important changes in how they manifest over time. Alternatively, it may reflect an actual, albeit slight, decline in psychopathology as people age. These patterns may not hold in clinical samples, where more severe symptoms could exacerbate maladaptive functioning, leading to a reciprocal worsening of both personality pathology and emotional disorders. This is in line with metaanalytic studies that show the median age at onset of GAD and depression at 30–35 years, and PD at 25–27 years⁶⁹. Nevertheless, our findings underscore the importance of considering both personality functioning and maladaptive personality traits in middle-aged and older individuals, as these factors may be crucial in the onset and progression of emotional disorders. Future longitudinal studies should further investigate the trajectory and interplay of symptomatic and personality disorders across the lifespan, contributing to the literature on which model best captures the associations between personality and psychopathology^6,8,10.

Clinical implications

Our findings further confirm the strong relationship between ED and personality pathology and therefore could have significant practical implications for clinical diagnosis and treatment planning. The study highlights the critical need for comprehensive assessment of personality pathology in all individuals seeking psychiatric help for symptoms typically associated with ED, such as anxiety and depression. This is particularly important given the well-established link between underlying personality pathology and more severe symptomatology, higher risk of treatment dropout, more frequent ruptures in the therapeutic alliance, poorer prognosis, and more severe clinical outcomes^70,71. It is likely that nonresponders to targeted ED treatments—such as cognitive-behavioral therapy (CBT) or supportive therapy—may have higher levels of undiagnosed personality pathology, which requires more intensive, structure-oriented psychotherapeutic interventions, such as schema therapy, mentalization-based treatment, transference-focused psychotherapy, or general psychiatric management (GPM)⁷². Moreover, early relational trauma may be a particularly relevant distal risk factor for both PD and ED, as personality functioning can act as either a risk or resilience factor in clinical outcomes related to childhood trauma⁵.

Our findings suggest that the five domains of maladaptive personality traits may aid clinicians in predicting the presence of both symptom disorders and personality pathology, thereby extending the clinical utility of the trait assessment. Considering the challenges related to the diagnostic accuracy of personality pathology in the presence of severe depressive symptoms—such as the susceptibility of self-reports to response bias due to depressed mood—attention should be paid to the type of diagnostic method used. To control for bias, the implementation of structured clinical interviews was advised⁵². Alternatively, integrating self-report data with clinical evaluations may enhance diagnostic precision, enabling meaningful attempts to explain discrepancies at an individual (case) level. Furthermore, given the need for comprehensive yet flexible personality assessment models that can guide the selection of measures⁷³, we suggest the diagnostic utility of an approach in which the assessment of depression and anxiety severity is accompanied by evaluations of personality disturbances as an indicator of severity and possible complications in positive treatment outcomes. This is in line with recommendations for multi-method assessment in diagnosing personality disorders^74,75, combining multiple sources of information enables a more holistic diagnostic perspective. Such an approach facilitates the consideration of the interplay between ego-dystonic symptoms (emotional disorders) and ego-syntonic symptoms (personality disorders), offering greater insight into the complex relationships between emotional and personality dysfunctions. Lastly, the present results suggest that the clinical utility of the SIFS and PiCD, may extend beyond assessing personality dysfunction and may also be valuable in the assessment of symptom disorders, similar to the previously established PID-5 ²⁷. In conclusion, identifying personality predictors of emotional disorders is crucial as it enables clinicians to tailor treatments to individual needs, identify high-risk patients early, enhance diagnostic accuracy, and deepen our understanding of the underlying pathology and its mechanisms. This insight is key to improving treatment outcomes and optimizing care for patients.

Limitations

Several aspects of the present study limit the robustness of our conclusions. Firstly, participants were recruited from an online platform, which may affect the representativeness of the sample, since the incentive-based participation model may introduce biases related to sample selection and motivation. Additionally, our reliance on questionnaire-based, self-report data may have introduced biases related to self-perception and social desirability; future studies should confirm these findings with structured interviews and clinical diagnoses. Moreover, we did not assess comorbidities between personality disorders and emotional disorders, which could have provided a more comprehensive understanding of the interplay between these conditions. The cross-sectional design further limits our ability to infer causal relationships between personality pathology and mental health outcomes seen as ED. While we adopted a predispositional model, which aligns with a specific causal model, this approach may overemphasize the idea that personality pathology leads to emotional disorders. However, contemporary perspectives on the etiology of mental disorders emphasize multiple etiological pathways with bidirectional influences (e.g⁷⁶). We cannot rule out alternative mechanisms, such as shared etiological factors (e.g., relational trauma) or overlapping diagnostic criteria⁶⁸. Furthermore, the generalizability of our findings to other populations (e.g., adolescents), diverse settings (e.g., primary care), and alternative diagnostic methods (e.g., categorical personality disorder diagnoses) remains uncertain.

Conclusions

This study contributes to the understanding of depression and anxiety symptoms by examining their relationships with personality dysfunction and pathological traits as defined in the DSM-5 Alternative Model for Personality Disorders¹³ and ICD-11¹². Consistent with our hypothesis, both personality dysfunction (identity and empathy) and pathological traits (Negative Affectivity and Detachment) emerged as significant predictors of these emotional disorders. These findings highlight the critical role of both components of the dimensional model of personality disorders in understanding depression and anxiety symptoms. They underscore the importance of screening for personality pathology in individuals presenting with these symptoms and considering it carefully in both diagnosis and treatment planning. Recognizing personality dysfunction alongside emotional disorders is essential for providing effective, personalized therapeutic interventions that address the underlying personality pathology driving these conditions.

Author contributions

M.J. and E.S. conceptualized and designed the study, supervised data collection, and conducted the primary data analysis. E.S. performed and described the final statistical analyses. M.J. drafted the manuscript, while E.S. provided critical revisions, and M.J. finalized the manuscript for submission. Both M.J. and E.S. prepared the tables. All authors reviewed and approved the final version of the manuscript.

Data availability

The datasets generated and analysed during the current study are available in the OSF repository: https://osf.io/nbt5s/?view_only=fd949fb62a9249fcb37d82d3cfa09fec.

Declarations

Competing interests

The authors declare no competing interests.